Code Review

---
name: code-review
version: 2.4.0
description: '[Code Quality] Use when evaluating review feedback, requesting a targeted code-quality review, or verifying completion claims.'
execution-mode: subagent
context-budget: critical
---

<!-- PROMPT-ENHANCE:STEP-TASK-ANCHOR:START -->

> **[BLOCKING]** Execute skill steps in declared order. NEVER skip, reorder, or merge steps without explicit user approval.
> **[BLOCKING]** Before each step or sub-skill call, update task tracking: set `in_progress` when step starts, set `completed` when step ends.
> **[BLOCKING]** Every completed/skipped step MUST include brief evidence or explicit skip reason.
> **[BLOCKING]** If Task tools are unavailable, create and maintain an equivalent step-by-step plan tracker with the same status transitions.

<!-- PROMPT-ENHANCE:STEP-TASK-ANCHOR:END -->

## Quick Summary

**Goal:** Ensure reviewed code is correct, easy to change, convention-aligned, and verification-backed before acceptance or handoff — via receiving feedback with verification (not performative agreement), requesting targeted systematic reviews through the code-reviewer subagent, and enforcing verification gates before completion claims.

**Summary:**

- **Purpose:** evaluate feedback, targeted review requests, or completion claims with a skeptical, evidence-first code-quality review; current-change or branch-diff requests route to `changes-review`.
- **Main steps (in order):** create the report → run graph/blast-radius and full-pipeline impact trace → detect risks, plan compliance, and file surfaces → review each file incrementally → run holistic and final assessments → validate findings, fix only validated findings that block the current round, and full re-review after fixes. Round 1 blocks on every validated severity; from round 2 onward CRITICAL/HIGH/MEDIUM block, LOW-only is deferred, and failed binary gates always block.
- **Evidence gates:** read relevant project references, cite every finding with `file:line`, trace consumers/dependents and tests, and use confidence thresholds before any acceptance claim.
- **Outcome:** a report-backed verdict that makes future change cheaper, with no performative agreement and no unvalidated fix or clean-pass claim.

> **Routing boundary:** If the user asks to review current changes, uncommitted work, staged/unstaged diffs, or a branch-to-branch diff, use `changes-review` instead.

> **Shared engine (keep in sync):** `code-review` and `changes-review` share the same review-protocol `SYNC:` blocks. Canonical source: `.claude/skills/shared/sync-inline-versions.md`; policy: `SYNC:shared-protocol-duplication-policy`. When you change a shared block in one skill, update the canonical file AND the sibling skill so the two never drift. The skills differ only in entry intent (explicit scope / feedback / completion-gate vs git diff) — not in review quality.

> **MANDATORY** Before reviewing, search for project-specific reference docs:
>
> **Coding standards** — search: `code-review-rules`, `coding-standards`, `style-guide`, `contributing`
> **Architecture** — search: `patterns-reference`, `architecture`, `adr`
> **Test conventions** — search: `integration-test-reference`, `test-guide`, `test-conventions`
> **Design system** — search: `design-system`, `design-tokens`, `component-library`
>
> Read found docs before reviewing. None found → rely on tech stack knowledge from file extensions/directory structure.

**Workflow:**

1. **Create Review Report** — Init `plans/reports/code-review-{date}-{slug}.md`
2. **Phase 0: Blast Radius** — Run graph analysis first if `.code-graph/graph.db` exists
3. **Phase 0.1: Change Context & Full-Pipeline Impact Trace (MANDATORY comprehension-first)** — Note the change context, then holistically trace the main affected area's full pipeline across BOTH boundaries — client↔server tier (FE↔BE) AND service/event/external — classifying each seam/touchpoint NONE/ADDITIVE/BREAKING (explicit N/A for single-tier or monolith)
4. **Phase 0.3: Risk Detection** — Detect dependency, migration, bus/event, API, security, config, and infra risks
5. **Phase 0.5: Plan Compliance** — Verify changed files and tests against active plan when present
6. **Phase 0.7: Surface Detection** — Classify files by language + directory semantics + change nature → route sub-agents; invoke `/ui-review` when frontend/UI files are present
7. **Phase 1: File-by-File** — Review each file, update report with correctness, convention, DRY, intent, test, and docs checks
8. **Phase 2: Holistic** — Re-read accumulated report, assess overall approach, architecture, duplication, and cross-boundary behavior
9. **Phase 3: Final Result** — Update report with overall assessment, critical issues, recommendations, docs staleness, and test gaps
10. **Fix Loop: Validate → Fix → Full Re-Review** — When findings exist, validate them first, fix only validated findings that block the current round, then restart the full review after the fix cycle; round-2+ LOW-only findings are recorded/deferred and do not open another round.

**Key Rules:**

- **Report-Driven**: Build report incrementally; re-read for big picture
- **Detect First**: Run graph blast radius when available, then classify change types and file surfaces before any review
- **Easy to Change for Code**: Treat future change cost as the primary code-quality metric; DRY, SOLID, abstraction, and patterns are tools only when they reduce change amplification
- **No Performative Agreement**: Technical evaluation only ("You're right!" banned)
- **Verification Gates**: Evidence required before completion claims
- **Review Current Diffs Elsewhere**: Current changes, staged/unstaged diffs, and branch diffs belong to `changes-review`
- **A clean review pass ENDS the review once the persisted `minRounds` is met.** Do not spend a fresh-context pass re-reviewing known findings before validation/fix; re-review after fixes change the target or to satisfy an explicitly declared independent-pass minimum.

# Code Review

Three practices: receiving feedback with technical rigor, requesting systematic reviews via code-reviewer subagent, enforcing verification gates before completion claims.

> Run `python .claude/scripts/code_graph query tests_for <function> --json` on changed functions to flag coverage gaps.

## Review Mindset (NON-NEGOTIABLE)

**Skeptical. Every claim needs traced proof `file:line`. Confidence >80% to act.**

- NEVER accept code correctness at face value — trace call paths
- NEVER include finding without `file:line` evidence (grep results, read confirmations)
- ALWAYS question: "Does this actually work?" → trace it. "Is this all?" → grep cross-service
- ALWAYS verify side effects: check consumers + dependents before approving

## First Principle — Easy to Change

> **The success metric of every coding decision is _future change cost_.**
> DRY, SRP, abstraction, design patterns, naming, layering, tests — every
> technique exists to serve one goal: **making the next change cheaper**.

When evaluating code, a refactor, a test, or an abstraction, ask:
**does this make the next change cheaper or more expensive?**

- Reject "best practices" that raise change cost (premature abstraction,
  speculative generality, leaky indirection, ceremony without payoff).
- Name the real enemies in findings: **coupling, hidden state, duplicated
  knowledge, unclear intent, irreversible decisions exposed too early**.
- Favor project-owned boundaries around external libraries, for example
  component/service input-output contracts, when they localize future library
  changes; reject pass-through wrappers that add ceremony without lowering
  change cost.
- A simpler design that is easy to change beats a sophisticated design that
  isn't.

Apply this lens **before** invoking any specific rule, pattern, or checklist
below — if a downstream rule would raise change cost, this principle wins.

---

## Core Principles (ENFORCE ALL)

| Principle          | Rule                                                                                                        |
| ------------------ | ----------------------------------------------------------------------------------------------------------- |
| **YAGNI**          | Flag code solving hypothetical problems (unused params, speculative interfaces)                             |
| **KISS**           | Flag unnecessary complexity. "Is there a simpler way?"                                                      |
| **DRY**            | Grep for similar/duplicate code. 3+ similar patterns → flag for extraction                                  |
| **Clean Code**     | Readable > clever. Names reveal intent. Functions do ONE thing. Nesting <=3. Methods <30 lines              |
| **Convention**     | MUST ATTENTION grep 3+ existing examples before flagging violations. Codebase convention wins over textbook |
| **No Bugs**        | Trace logic paths. Verify edge cases (null, empty, boundary). Check error handling                          |
| **Proof Required** | Every claim backed by `file:line` evidence. Speculation is forbidden                                        |
| **Doc Staleness**  | Cross-ref changed files against related docs. Flag stale/missing updates                                    |

**Technical correctness over social comfort.** Verify before implementing. Evidence before claims.

## Graph-Enhanced Review (RECOMMENDED if graph.db exists)

1. `python .claude/scripts/code_graph graph-blast-radius --json` — prioritize files by impact (most dependents first)
2. `python .claude/scripts/code_graph query tests_for <function_name> --json` — flag untested changed functions
3. `python .claude/scripts/code_graph trace <file> --direction downstream --json` — downstream impact (events, bus, cross-service)
4. `python .claude/scripts/code_graph trace <file> --direction both --json` — full flow context for controllers/commands/handlers
5. Wide blast radius (>20 impacted nodes) = high-risk. Flag in report.

## Review Approach (Report-Driven Two-Phase — CRITICAL)

**MANDATORY FIRST: Create Todo Tasks**

| Task                                                                   | Status      |
| ---------------------------------------------------------------------- | ----------- |
| `[Review] Create report file`                                          | in_progress |
| `[Review Phase 0] Run graph blast-radius if available`                 | pending     |
| `[Review Phase 0.1] Note change context + holistic full-pipeline trace across BOTH boundaries — client↔server tier (FE↔BE) AND service/event/external — classify each seam/touchpoint NONE/ADDITIVE/BREAKING (MANDATORY comprehension-first; N/A for single-tier/monolith)` | pending     |
| `[Review Phase 0.3] Detect high-risk change types`                     | pending     |
| `[Review Phase 0.5] Plan compliance check (skip if no active plan)`    | pending     |
| `[Review Phase 0.7] Detect categories + route sub-agents`              | pending     |
| `[Review Phase 0.7b] /ui-review sub-review — skip if no frontend/UI files in changeset` | pending     |
| `[Review Phase 1] File-by-file review + update report`                 | pending     |
| `[Review Phase 2] Holistic assessment`                                 | pending     |
| `[Review Phase 3] Final findings, docs triage, and test sync findings` | pending     |
| `[Review Fix Loop] Validate findings, fix validated blocking findings, and full re-review until the current severity bar is clear` | pending     |
| `[Review Final] Consolidate all rounds`                                | pending     |

**Step 0: Create Report File**

Create `plans/reports/code-review-{date}-{slug}.md` with Scope, Files to Review sections.

**Phase 0: Graph Blast Radius (FIRST WHEN AVAILABLE)**

If `.code-graph/graph.db` exists, run graph impact analysis before reviewing:

- `python .claude/scripts/code_graph graph-blast-radius --json` or the project equivalent
- Record impacted files count, untested changed functions, and risk level in the report
- Prioritize high-impact files during Phase 1

If graph data is unavailable, record "Graph not available — skipping blast radius" and proceed to Phase 0.1.

**Phase 0.1: Change Context Comprehension & Full-Pipeline Impact Trace (MANDATORY — comprehension-first)**

> **IMPORTANT MANDATORY MUST ATTENTION:** First *comprehension* step — before any file-by-file or dimensional review. Blast radius (Phase 0) gathers impact data; here holistically UNDERSTAND the change, trace main affected area's full pipeline across every boundary it crosses. Apply BOTH the **Cross-Stack Impact Trace** and **Cross-Service Check** protocols (bodies in the SYNC section below). This holistic first-pass map FEEDS the later Phase 0.3 change-type risk tasks and the Phase 2 holistic cross-boundary assessment — does not replace them.

Write a one-paragraph **Change Context** note (what changed · intent · originating tier · main affected feature/flow), then run both traces per the **Cross-Stack Impact Trace** and **Cross-Service Check** protocols (SYNC blocks below).

**Phase 0.3: Detect High-Risk Change Types**

Before file review, inspect the target diff or explicit file set for:

- Bugfix, failed verification, stale/incorrect final output, regression, or behavior-changing fix — require `Debugger Trace: End -> Start`, all feeder paths, hypothesis matrix, owning fix layer, and forward convergence proof; missing trace evidence is a High/Critical review finding
- Dependency upgrades — semver, breaking changes, advisories, peer compatibility
- Migrations or schema changes — rollback, lock/volume impact, zero-downtime deployment, idempotent backfill
- Bus events/messages — consumer existence, idempotency, retries, poison/dead-letter handling
- API contract changes — backward compatibility, caller alignment, auth, required response fields
- Security changes — enforcement coverage, privilege escalation, negative tests, duplicated permission strings
- Config/env changes — all environments covered, no secrets, fail-fast behavior, setup docs
- Infra changes — dev/prod parity, pinned versions, CI/CD permissions, reproducible builds

Create focused review tasks for every true signal and complete them before dimensional review.

**Phase 0.5: Plan Compliance Check (CONDITIONAL)**

If active plan context exists, verify scope, test evidence, and success criteria against the plan before file review; otherwise record the skip reason.

**Goal Contract mapping (CONDITIONAL — when an active goal exists):** Resolve the active Goal Contract per the goal-contract-satisfaction-loop protocol (active plan `goal.md` → `plans/goals/{YYMMDD-HHmm}-{slug}/goal.md`). When found, map the reviewed changes to the saved success criteria in the report — which criteria this changeset advances (with `file:line` evidence), which it leaves untouched, and any change serving NO saved criterion (flag as scope drift unless justified). Record `No active goal — mapping skipped.` when none exists; do NOT create a goal file from inside a review.

**Phase 0.7: Detect Review Categories**

Before any review — classify the changeset and route sub-agents:

| Signal in changed files                  | Route to                                                |
| ---------------------------------------- | ------------------------------------------------------- |
| Auth/permission/token/encryption files   | `security-auditor`                                      |
| Query files, caching, batch processing   | `performance-optimizer`                                 |
| Source code (logic, handlers, services)  | `code-reviewer`                                         |
| Frontend/UI files (components, templates, `.html`/`.scss`/`.css`, design-system) | `/ui-review` skill (see Phase 0.7b) |
| Docs, plans, specs, markdown             | `general-purpose`                                       |
| Mixed changeset with security/perf files | Spawn specialized sub-agent first, then `code-reviewer` |

**Phase 0.7b: Frontend/UI Sub-Review (CONDITIONAL — `/ui-review`)**

If the changeset contains any frontend/UI files matching the project's configured UI patterns (components, templates, `.html`/`.scss`/`.css`, design-system tokens), invoke the `/ui-review` skill as a sub-review so UI-specific concerns are covered — long-content overflow (wrap vs ellipsis+tooltip), responsive multi-screen flex, flex-grow with min/max over fixed px, semantic z-index discipline (no raw numbers, no `!important`), and BEM classes on all template elements. Fold its findings into this report's Phase 3 results.

**Skip (record reason)** when no frontend/UI files are present in the changeset — log "Skipped Phase 0.7b — no frontend/UI files in changeset".

**Phase 0.8: Derive Review Categories**

Group changed files by: file language (extension), directory semantics (path), change nature (new entity, schema, config, UI, test).

For each category: name it, create sub-task, derive concerns using `SYNC:category-review-thinking` (first principles — NOT a fixed checklist).

> Category list = Phase 1 work breakdown. Each category → own section in report.

**Phase 1: File-by-File Review (Build Report)**

For EACH file, immediately update report:

- File path, Change Summary, Purpose, Issues Found
- **Convention check:** Grep 3+ similar patterns — does new code follow existing convention?
- **Correctness check:** Trace logic — null, empty, boundary, error cases handled?
- **DRY check:** Grep for similar/duplicate code — does this logic exist elsewhere?
- **Intention check:** Does the change serve the stated purpose? Flag unrelated modifications
- **Test check:** Changed behavior has corresponding test/spec coverage or a documented gap
- **Documentation check:** Related docs/specs/READMEs still match the changed behavior

**Phase 2: Holistic Review (Re-read Report)**

After all files reviewed, re-read accumulated report:

- **Technical Solution**: Overall approach coherent as unified plan?
- **Responsibility**: Logic in LOWEST layer? Business logic not in controllers?
- **Data ownership**: Constants/config in model/entity, not controller/component?
- **Duplication**: Grep to verify — duplicated logic across changes?
- **Architecture**: Clean Architecture? Service boundaries respected?
- **Plan Compliance**: If active plan → check `## Plan Context`: impl matches requirements, TCs have code evidence (not "TBD"), no requirement unaddressed
- **Design Patterns**: Pattern opportunities (switch→Strategy)? Anti-patterns (God Object, Copy-Paste, Circular Dep)? DRY via base classes?
- **Cross-Boundary Behavior**: Callers/callees aligned? API/event contracts consistent? New wiring reachable?
- **Test Sync**: Business logic changes have corresponding tests or explicit user-facing gap
- **Translation Sync**: Multilingual UI text changes have translation updates or explicit risk acceptance
- **Bugfix Trace Completeness**: If the diff is a bugfix or behavior-changing fix, the review report must state whether final-state trace, feeder paths, hypothesis matrix, owning fix layer, forward convergence proof, and tests/proof mapping are complete

**MUST ATTENTION CHECK — Spec-Loop Test Discipline (changed core logic):** Beyond the happy/error path traces above, hold changed core logic to a hard-to-fake bar. Apply a **MUTATION-SCORE** bar — a surviving mutant means a missing invariant, so demand the killing test — and do NOT accept a line-coverage % as proof of test strength. Flag any `[HARD]`/§5 invariant whose only coverage is example tests with no universally-quantified **property TC** (plus boundary counter-case) as a HIGH finding. Every behavior-changing finding requires a **Dual-Feedback row** (does it feed the spec? does it feed the tests? a blank axis = INCOMPLETE) — record it in the report. Adjudicate any spec divergence per `SYNC:spec-drift-adjudication` (CODE-WRONG / SPEC-STALE / AMBIGUOUS / **SPEC-SILENT**); a **SPEC-SILENT** finding — the code correctly enforces an invariant NO spec artifact states — has BOTH axes non-N/A: Spec feedback = add the missing §4 BR / §3 AC (+ §5 invariant if applicable) and a §8 TC via `/spec [update]` + `/spec [mode=tests]`; Test feedback = the new property/regression test guarding the now-written invariant — never leave a discovered invariant only in code or only in tests. Review the **whole package** (spec + tests + code), not just the diff, so the spec is enriched, not just patched.

**MUST ATTENTION CHECK — Clean Code:** YAGNI (unused params, speculative interfaces)? KISS (simpler exists)? Methods >30 lines or nesting >3?

**MUST ATTENTION CHECK — Correctness:** Null/empty/boundary handled? Error paths caught? Async race conditions? Trace happy + error paths.

**Documentation Staleness Check:**

For each changed file — grep file name/module across `docs/` and AI tooling dirs. Changed behavior → flag stale doc (specific section + what changed). **Flag the staleness only — never auto-fix docs here.**

Common staleness patterns: count/limit changed → docs embedding that number | API/contract changed → API usage docs | hook/skill added/removed → catalogs/README | schema changed → entity reference docs.

**Phase 3: Final Review Result**

Update report: Overall Assessment, Critical Issues, High Priority, Architecture Recommendations, Cross-Boundary Impact (from Phase 0.1 — per client↔server seam AND per service/event/external touchpoint: NONE / ADDITIVE / BREAKING with routed fix; or explicit "Single-tier / monolith — N/A"), Documentation Staleness, Positive Observations.

If documentation staleness is detected, recommend `docs-update` and list exact stale sections; do not silently pass stale docs.

## Validated Fix + Full Re-Review (MANDATORY when findings are fixed)

After Phase 3, do not spawn a fresh reviewer just to re-review the same finding set. First validate findings, then fix only validated findings that block the current round. Because fixes change the review target, restart the full review after the fix cycle; from round 2 onward a LOW-only result is deferred and does not trigger another cycle. If that restarted protocol uses sub-agents, construct each Agent call with the canonical template from `SYNC:review-protocol-injection`:

1. Copy Agent call shape from `SYNC:review-protocol-injection` verbatim
2. Embed full verbatim body of all 11 SYNC blocks: `SYNC:spec-tests-code-triangulation`, `SYNC:evidence-based-reasoning`, `SYNC:bug-detection`, `SYNC:design-patterns-quality`, `SYNC:complexity-prevention`, `SYNC:logic-and-intention-review`, `SYNC:test-spec-verification`, `SYNC:fix-layer-accountability`, `SYNC:rationalization-prevention`, `SYNC:graph-assisted-investigation`, `SYNC:understand-code-first`
3. Task: `"Run a full fresh code-review pass over the current assigned scope after validated fixes were applied. Focus: cross-cutting concerns, interaction bugs, convention drift, missing pieces, subtle edge cases, logic errors, test spec gaps, and regressions introduced by the fixes."`
4. Target Files: `"use the explicit files, plan scope, or reviewer-provided target range"`
5. Report: `plans/reports/code-review-rerun{N}-{date}.md`

After sub-agent returns:

1. **Read** report from `plans/reports/code-review-rerun{N}-{date}.md`
2. **Integrate** findings as `## Re-Review {N} Findings` — DO NOT filter or override
3. **If findings remain:** validate the new finding set before any additional fixes
4. **Repeat only after another fix cycle:** restart the full review again after validated fixes are applied; if the same blocker repeats across 2 full invocations with no progress, escalate via `AskUserQuestion`

## Clean Code Rules (MUST ATTENTION CHECK)

| #   | Rule                      | Details                                                                                                                                 |
| --- | ------------------------- | --------------------------------------------------------------------------------------------------------------------------------------- |
| 1   | **No Magic Values**       | All literals → named constants                                                                                                          |
| 2   | **Type Annotations**      | Explicit parameter and return types on all functions                                                                                    |
| 3   | **Single Responsibility** | One concern per method/class. Event handlers/consumers: one handler = one concern. NEVER bundle — a framework event dispatcher can swallow handler exceptions silently |
| 4   | **DRY**                   | No duplication; extract shared logic                                                                                                    |
| 5   | **Naming**                | Names reveal purpose, not mechanism: specific names (`orderRecords` not `data`), Verb+Noun methods, is/has/can/should booleans, no unexplained abbreviations; public/cross-layer abstractions name the capability or domain contract and keep provider details on concrete adapters (`IStorage`/`Storage` → `AzureBlobStorage`), with narrow contracts and local interface syntax |
| 6   | **Performance**           | No O(n²) (use dictionary). Project in query (not load-all). ALWAYS paginate. Batch-by-IDs (not N+1)                                     |
| 7   | **Entity Indexes**        | Collections: index management methods. EF Core: composite indexes. Expression fields match index order. Text search → text indexes      |

## Data Lifecycle Rules (MUST ATTENTION CHECK)

**Decision test:** _"Delete the DB and start fresh — does this data still need to exist?"_ Yes → **Seeder/fixture**. No → **Migration**.

| Type                 | Contains                                                                                | NEVER contains                                   |
| -------------------- | --------------------------------------------------------------------------------------- | ------------------------------------------------ |
| **Seeder / Fixture** | Default records, system config, reference data (idempotent — safe to run every startup) | Schema changes                                   |
| **Migration**        | Schema changes, column adds/removes, data transforms, index changes                     | Default records, permission seeds, system config |

Apply project's language/framework conventions. Principle universal — implementation project-specific.

## Legacy Pattern Compliance

When reviewing files with legacy and modern patterns:

1. **Detect legacy signals** — search `project-config.json`, `package.json`, or equivalent for `"legacy"`, version flags, feature annotations
2. **Read what "legacy" means** — grep 3+ legacy files to understand pattern constraints vs. modern files
3. **Derive compliance rules** — what lifecycle/memory management differences exist between legacy/modern for this tech stack?
4. **Apply tech stack knowledge** to flag anti-patterns

NEVER assume any specific framework's lifecycle. Derive from codebase evidence.

## When to Use This Skill

| Practice               | Triggers                                                                                   | MUST ATTENTION READ                            |
| ---------------------- | ------------------------------------------------------------------------------------------ | ---------------------------------------------- |
| **Receiving Feedback** | Review comments received, feedback unclear/questionable, conflicts with existing decisions | `references/code-review-reception.md`          |
| **Requesting Review**  | After each subagent task, major feature done, targeted review scope, after complex bug fix | `references/requesting-code-review.md`         |
| **Verification Gates** | Before any completion claim, commit, push, or PR. ANY success/satisfaction statement       | `references/verification-before-completion.md` |

## Quick Decision Tree

```
SITUATION?
│
├─ Received feedback
│  ├─ Unclear items? → STOP, ask for clarification first
│  ├─ From human partner? → Understand, then implement
│  └─ From external reviewer? → Verify technically before implementing
│
├─ Completed work
│  ├─ Major feature/task? → Request code-reviewer subagent review
│  └─ Before merge? → Request code-reviewer subagent review
│
└─ About to claim status
   ├─ Have fresh verification? → State claim WITH evidence
   └─ No fresh verification? → RUN verification command first
```

## Receiving Feedback Protocol

**Pattern:** READ → UNDERSTAND → VERIFY → EVALUATE → RESPOND → IMPLEMENT

- NEVER use performative agreement ("You're right!", "Great point!", "Thanks for...")
- NEVER implement before verification
- MUST ATTENTION restate requirement, ask questions, or push back with technical reasoning
- ask for clarification on ALL unclear items BEFORE starting
- grep for usage before implementing suggested "proper" features (YAGNI check)

**Source handling:** Human partner → implement after understanding. External reviewer → verify technically, push back if wrong.

**Full protocol:** `references/code-review-reception.md`

## Requesting Review Protocol

1. Get git SHAs: `BASE_SHA=$(git rev-parse HEAD~1)` and `HEAD_SHA=$(git rev-parse HEAD)`
2. Dispatch code-reviewer subagent with: WHAT_WAS_IMPLEMENTED, PLAN_OR_REQUIREMENTS, BASE_SHA, HEAD_SHA, DESCRIPTION
3. Act on feedback: Critical → fix immediately. Important → fix before proceeding. Minor → note for later.

**Full protocol:** `references/requesting-code-review.md`

## Verification Gates Protocol

**Iron Law: NO COMPLETION CLAIMS WITHOUT FRESH VERIFICATION EVIDENCE**

**Gate:** IDENTIFY command → RUN it → READ output → VERIFY it confirms claim → THEN claim. Skip any step = lying.

| Claim            | Required Evidence               |
| ---------------- | ------------------------------- |
| Tests pass       | Test output shows 0 failures    |
| Build succeeds   | Build command exit 0            |
| Bug fixed        | Original symptom test passes    |
| Requirements met | Line-by-line checklist verified |

**Red Flags — STOP:** "should"/"probably"/"seems to", satisfaction before verification, committing without verification, trusting agent reports.

**Full protocol:** `references/verification-before-completion.md`

## Related

- `code-simplifier`
- `debug-investigate`
- `refactoring`

---

## Systematic Review Protocol (10+ changed files)

> When Phase 1 finds 10+ changed files, apply the **Systematic Review Batching** protocol (map-reduce: size-capped batches + hierarchical synthesis) defined below.

---

## Workflow Recommendation

> **MANDATORY — NO EXCEPTIONS:** If NOT already in a workflow, use `AskUserQuestion` to ask user:
>
> 1. **Activate `workflow-review-changes` workflow** (Recommended) — full review → validated current-round blocking fix cycle → re-review until the severity bar is clear (Round 2+ LOW-only findings are deferred)
> 2. **Execute `/code-review` directly** — run standalone

---

## Architecture Boundary Check

For each changed file, verify no forbidden layer imports:

1. **Read rules** from `docs/project-config.json` → `architectureRules.layerBoundaries`
2. **Determine layer** — match file path against each rule's `paths` glob patterns
3. **Scan imports** — grep for the configured language's import/include statements
4. **Check violations** — import path contains forbidden layer name → violation
5. **Exclude framework** — skip files matching `architectureRules.excludePatterns`
6. **BLOCK on violation** — `"BLOCKED: {layer} layer file {filePath} imports from {forbiddenLayer} ({importStatement})"`

If `architectureRules` absent in project-config.json → skip silently.

---

## Phase 4: Why-Review Self-Validation Gate (MANDATORY when findings exist)

> **Purpose:** Adversarial validation of own findings BEFORE handoff. Catches over-flagged Highs, false positives, and severity inflation at the source rather than letting them propagate downstream.

**Trigger:** Any finding produced (Critical, High, Medium, OR Low). Skip ONLY when the report's verdict is unconditional PASS with literally zero findings.

**Protocol:**

1. Read own finalized report from `plans/reports/{skill}-{date}-{slug}.md`
2. Invoke `/why-review` skill with arg: `validate findings in plans/reports/{skill}-{date}-{slug}.md — verify each finding has file:line proof, steel-man each rejected interpretation, and stress-test severity classifications`
3. Read the validation verdict path returned by why-review, expected as `plans/reports/why-review-validate-{date}.md`
4. **If why-review demotes/removes any finding:** UPDATE own finalized report with revised severities, remove false positives, and add a `## Why-Review Validation Notes` section citing what changed and why
5. **If why-review confirms all findings:** Append `## Why-Review Validation` line to own report stating "All N findings re-validated against actual code; no severity changes."

**Skip conditions (record explicit reason if skipping):**

- Verdict is unconditional PASS with zero findings → log "Skipped — no findings to validate"
- Why-review skill itself is the active context (avoid recursion)

**Why this exists:** AI sub-agent reports inherit confirmation bias — the orchestrator absorbs severity claims as ground truth. The 2026-05-09 review incident produced 5 Highs; adversarial validation demoted 3 of them. Codify this as standard practice.

---

## Next Steps

**MANDATORY — NO EXCEPTIONS** after completing, use `AskUserQuestion`:

- **"/fix (Recommended)"** — review found issues needing fixes
- **"/watzup"** — review clean, wrap up session
- **"Skip, continue manually"** — user decides

## AI Agent Integrity Gate (NON-NEGOTIABLE)

**Completion ≠ Correctness.** Before reporting ANY work done:

1. **Grep every removed name.** Extraction/rename/delete → grep confirms 0 dangling refs across ALL file types.
2. **Ask WHY before changing.** Existing values intentional until proven otherwise.
3. **Verify ALL outputs.** One build passing ≠ all builds passing.
4. **Evaluate pattern fit.** Copying nearby code? Verify preconditions match — scope, lifetime, base class, constraints.
5. **New artifact = wired artifact.** Created something? Prove it's registered, imported, reachable by all consumers.

---

> **[IMPORTANT]** Use `TaskCreate` to break ALL work into small tasks BEFORE starting — including tasks for each file read. This prevents context loss from long files. For simple tasks, AI MUST ATTENTION ask user whether to skip.

> **Critical Purpose:** Ensure quality — no flaws, bugs, missing updates, stale content. Verify code AND documentation.

> **External Memory:** Complex work → write findings incrementally to `plans/reports/` — prevents context loss, serves as deliverable.

> **Evidence Gate:** MANDATORY — every claim, finding, recommendation requires `file:line` proof + confidence % (>80% act, <80% verify first).

> **OOP & DRY:** MANDATORY — flag patterns extractable to base class/generic/helper. Same-suffix/lifecycle/responsibility classes share common base. Apply idiomatic abstraction (base class, mixin, trait, protocol) for project's language. Verify linting/analyzer configured.

<!-- SYNC:cross-stack-impact-trace -->

> **Cross-Stack Impact Trace** — FIRST review action: comprehend change holistically, THEN judge files. Every reviewed diff: note change context, trace full pipeline of main affected area end-to-end across client↔server seam, so a change on one tier can never silently break the other. (Distinct from `SYNC:cross-service-check`, which owns service-to-service / event boundary — this owns client↔server tier seam inside one app; pair both for full-pipeline coverage.)
>
> 1. **Comprehend context FIRST** — before file-by-file review, write short **Change Context** note: what changed, intent (why), originating tier (frontend / backend / shared / infra), main affected feature/flow. Do before flagging anything.
> 2. **Identify cross-stack seam(s)** — for main affected area, locate contract seam(s) between client and server: API route/endpoint + verb, request/response DTO or payload shape, shared type/schema, event/message contract, query/route params. Infer tier layout from `docs/project-config.json` and project conventions.
> 3. **Trace full pipeline end-to-end, in change's direction:**
>     - **Backend change → trace FORWARD to every frontend consumer:** handler/controller → response DTO/serializer → API client/service → store/state → component/template rendering or submitting it.
>     - **Frontend change → trace BACKWARD to backend contract:** component/form → API client call → route/endpoint → request DTO/validation → handler/domain.
>     - When `.code-graph/graph.db` exists, use `/graph-connect-api` and `python .claude/scripts/code_graph trace <file> --direction both --json` to map connection; otherwise grep route path, DTO/type name, each field name across BOTH tiers.
> 4. **Verify BOTH sides still agree** — for every changed seam confirm other tier matches: route path & verb, field names & types, nullability/optionality, required vs optional params, enum values, auth/permission, error/status shape. Any mismatch = **BREAKING** finding (backend change breaks a frontend consumer, or frontend now sends what backend rejects).
> 5. **Classify each seam:** NONE (no contract change) / ADDITIVE (backward-compatible) / BREAKING (consumer on other tier must change too). BREAKING seam whose other-tier consumer NOT updated in same diff = HIGH severity minimum (CRITICAL for auth/money/data-integrity paths).
>
> **Skip ONLY** when change has no cross-tier seam — pure docs, pure styling with no data contract, or single-tier tooling. State explicitly: `Single-tier change — no cross-stack seam`. Backend-only or single-tier repo still traces internal consumers (`SYNC:cross-service-check` for service/event boundaries).
>
> **BLOCKED until:** Change Context noted · seam(s) identified or explicit N/A · full pipeline traced in change direction · every changed seam classified NONE / ADDITIVE / BREAKING.

<!-- /SYNC:cross-stack-impact-trace -->

<!-- SYNC:cross-service-check -->

> **Cross-Service Check** — Microservices/event-driven: MANDATORY before concluding investigation, plan, spec, or feature doc. Missing downstream consumer = silent regression.
>
> | Boundary            | Grep terms                                                                      |
> | ------------------- | ------------------------------------------------------------------------------- |
> | Event producers     | `Publish`, `Dispatch`, `Send`, `emit`, `EventBus`, `outbox`, `IntegrationEvent` |
> | Event consumers     | `Consumer`, `EventHandler`, `Subscribe`, `@EventListener`, `inbox`              |
> | Sagas/orchestration | `Saga`, `ProcessManager`, `Choreography`, `Workflow`, `Orchestrator`            |
> | Sync service calls  | HTTP/gRPC calls to/from other services                                          |
> | Shared contracts    | OpenAPI spec, proto, shared DTO — flag breaking changes                         |
> | Data ownership      | Other service reads/writes same table/collection → Shared-DB anti-pattern       |
>
> **Per touchpoint:** owner service · message name · consumers · risk (NONE / ADDITIVE / BREAKING).
>
> **BLOCKED until:** Producers scanned · Consumers scanned · Sagas checked · Contracts reviewed · Breaking-change risk flagged

<!-- /SYNC:cross-service-check -->

<!-- SYNC:systematic-review-batching -->

> **Systematic Review Batching (map-reduce)** — When a changeset is large, do NOT review files one-by-one. Partition into size-capped batches, fire one specialized sub-agent per batch in parallel, then reduce. This bounds EVERY context — each batch agent AND the orchestrator — so coverage stays complete as file count grows.
>
> **Trigger ladder (one ordered escalation — not competing thresholds):**
>
> 1. **< 10 changed files** → sequential per-file review (default; no batching).
> 2. **≥ 10 changed files** → switch to systematic parallel mode. Announce: `"Detected {N} changed files. Switching to systematic parallel review protocol."` Then: categorize → size-capped batches → flat consolidation.
> 3. **categories > 6 OR files > 40** → additionally insert the hierarchical synthesis tier (below). Everything from rung 2 still applies.
>
> **Step 1 — Categorize.** Group changed files into logical categories derived from the project's actual structure (not forced). Category is the *concern axis*; orient with these examples, derive what fits the repository:
>
> | Category Type | Example Groupings |
> | --- | --- |
> | Agent/Tooling | AI scripts, hooks, skill definitions, workflow configs, linting rules |
> | Root config/docs | Root README, project config, CI/CD pipeline configs |
> | Reference docs | Architecture docs, patterns references, setup guides |
> | Feature/domain docs | Business feature documentation, spec files, ADRs |
> | Backend logic | Service/handler/controller source (infer from project structure) |
> | Frontend logic | UI component/state/API source (infer from project structure) |
> | Data/Schema | Migrations, schema files, seed data |
> | Tests | Unit, integration, E2E test files |
> | Infrastructure | Docker, k8s, CI/CD, cloud manifests |
>
> **Step 2 — Size-capped batches.** One sub-agent per batch of **≤8 files OR ≤2000 diff-lines**, whichever hits first. Category stays the concern axis, but any category exceeding a cap splits into multiple size-capped batches (30 backend files → 4 batches). Size caps — not category caps — make "many files" safe: a category cap alone lets one giant category blow a single agent's context.
>
> **Step 2a — Sub-agent type per batch** (match the batch's dominant concern):
>
> - Code logic (any stack) → `code-reviewer`
> - Security-sensitive changes → `security-auditor`
> - Performance-critical paths → `performance-optimizer`
> - Docs, plans, specs, configs, infra → `general-purpose`
>
> Each batch sub-agent receives: its full file list; `SYNC:category-review-thinking` as its primary thinking model — derive each category's concerns from first principles, NOT a fixed checklist (if the consuming skill does not carry that block, apply category-first thinking directly); project reference docs relevant to its concern (discover via `*patterns*`, `*conventions*`, `*style-guide*`); cross-reference verification instructions (counts, tables, links). All batch agents run in parallel and write findings to `plans/reports/` (per `SYNC:task-tracking-external-report`); reducers read from disk, never from memory.
>
> **Step 3 — Reduce.**
>
> - **Flat reduction (rung 2, ≤6 categories AND ≤40 files):** the orchestrator collects each batch report, cross-references counts/tables/contracts ACROSS batches, detects gaps visible only across categories (feature in code but missing from docs; new API endpoint with no client call), and consolidates into one categorized holistic report.
> - **Hierarchical reduction (rung 3, > 6 categories OR > 40 files):** insert a mid-tier — each concern gets ONE synthesizer agent that reads only its own batch reports and emits a single concern-synthesis. The orchestrator reads the **concern-syntheses (~5)**, never the raw batch reports — keeping the reducer's context O(#concerns), not O(#files).
>   - **Cross-concern interaction pass (mandatory at rung 3 — closes the synthesis-tier blind spot):** concern-siloed synthesis can drop an interaction spanning two concerns AND two batches (tainted source in data-layer/batch 7 → sink in api/batch 3). So: (a) each concern-synthesizer MUST emit an explicit **"cross-concern interaction candidates"** list — entities/symbols/contracts it touched that plausibly bind to another concern (shared DTOs, event names, table/collection names, exported symbols); (b) the orchestrator MUST run the Step-3 cross-reference/gap step **over those candidate lists across all concern-syntheses**, not only within a batch, before concluding. Without this pass the tier trades completeness for context-bounding on exactly the large diffs it targets.
>
> **Step 4 — Holistic assessment.** With all findings combined, judge: overall coherence as a unified intent; cross-category sync (docs match code? contracts match callers?); risk areas where categories interact; missing doc/spec updates for changed artifacts.
>
> **No silent truncation.** If any cap forces sampling or a batch is dropped for budget, ANNOUNCE the dropped/sampled scope explicitly — bounded coverage must never read as complete coverage.

<!-- /SYNC:systematic-review-batching -->

<!-- SYNC:end-to-start-debugger-trace -->

> **End-to-Start Debugger Trace** — For non-trivial bugs, failed verification, regression fixes, behavior-changing code, or unclear code flow, start from the observed final state and walk backward before proposing a fix.
>
> 1. **Frame 0: observed end state** — Name the exact user-visible output, failing assertion, log line, persisted value, API response, rendered UI, or aggregate bucket. Record the reader/query/renderer that produced it with `file:line` evidence.
> 2. **Walk backward one hop at a time** — Trace final reader -> projection/cache/storage -> writer -> consumer/handler/job -> producer/caller -> original trigger. At every hop record: input, transformation, output, owner, and evidence.
> 3. **Enumerate all feeder paths** — Find every upstream producer/caller/event/job that can write into the final path, including retry, async, cache, background, and alternate UI/API paths. Mark each path verified, ruled out, or still unknown.
> 4. **Build the hypothesis matrix** — For each plausible cause, list evidence for, evidence against, how to reproduce/verify, blast radius, and status (`primary`, `contributing`, `ruled out`, `latent`). Do not fix until competing causes are explicitly resolved or bounded.
> 5. **Choose the owning fix layer** — Identify the invariant owner and the lowest shared point that protects all downstream consumers. A fix at the symptom site is rejected unless the symptom site owns the invariant.
> 6. **Prove convergence forward** — After choosing the fix, walk start -> end again and show how the corrected state reaches the observed final output. Map each root cause to a fix part and each fix part to a test/proof.
>
> **BLOCKED until:** final state named · backward trace written · all feeder paths enumerated · hypothesis matrix completed · owning fix layer justified · forward convergence proof mapped to tests.
>
> **NEVER:** Start at the first suspicious code path. Collapse multiple producers into one "flow". Treat duplicate symptoms as duplicate records without proving the read model. Skip ruled-out hypotheses.

<!-- /SYNC:end-to-start-debugger-trace -->

<!-- SYNC:graph-assisted-investigation -->

> **Graph-Assisted Investigation** — MANDATORY when `.code-graph/graph.db` exists.
>
> **HARD-GATE:** MUST ATTENTION run at least ONE graph command on key files before concluding any investigation.
>
> **Pattern:** Grep finds files → `trace --direction both` reveals full system flow → Grep verifies details
>
> | Task                | Minimum Graph Action                         |
> | ------------------- | -------------------------------------------- |
> | Investigation | `trace --direction both` on 2-3 entry files  |
> | Fix/Debug           | `callers_of` on buggy function + `tests_for` |
> | Feature/Enhancement | `connections` on files to be modified        |
> | Code Review         | `tests_for` on changed functions             |
> | Blast Radius        | `trace --direction downstream`               |
>
> **CLI:** `python .claude/scripts/code_graph {command} --json`. Use `--node-mode file` first (10-30x less noise), then `--node-mode function` for detail.

<!-- /SYNC:graph-assisted-investigation -->

<!-- SYNC:category-review-thinking -->

> **Category Review Thinking** — A thinking framework for reviewing any category of changed files. NOT a fixed checklist — derive concerns from domain knowledge; the examples are starting points only. Your knowledge of the category exceeds any list here — trust it.
>
> **Step 1 — Understand the category's role.** What is this category responsible for in the overall system? What invariants must it uphold? What are its consumer contracts (who depends on it, what do they expect)?
>
> **Step 2 — Read project conventions for this category.** Search for reference docs, style guides, ADRs, or READMEs specific to this area. Grep 3+ existing similar files — extract naming conventions, structural patterns, shared base classes. If no docs exist, derive conventions empirically from existing code.
>
> **Step 3 — Derive concerns from first principles.** Apply all that are relevant; expand beyond this list based on the actual category:
>
> - **Correctness:** Does the logic match the intent? Trace happy path AND error path.
> - **Boundary contracts:** Are interfaces/APIs/events/protocols honored? No implicit coupling introduced?
> - **Project conventions:** Does new code follow the patterns found in Step 2? Evidence-confirmed, not assumed.
> - **Security:** Auth enforced at every entry point? Input validated at boundaries? No secrets in the diff?
> - **Performance:** Unbounded operations? N+1 patterns? Blocking calls in async context? Unindexed queries?
> - **Maintainability:** DRY? Single responsibility? Complexity within reason? Names reveal intent?
> - **Boundary naming:** When the category exposes public or cross-layer types, APIs, events, or modules, verify that names describe the capability, domain purpose, or contract rather than the current provider/framework/transport; concrete adapters may carry those details. Check callers and implementations before flagging a name, and treat generic names (`Manager`, `Helper`, `Utils`, `Data`) as signals rather than automatic violations.
> - **Test coverage:** Are the changed paths covered by tests? Are existing tests still valid after the change?
> - **Documentation:** Do related docs, specs, or READMEs reflect the changes?
>
> **Step 4 — Create sub-tasks and execute.** For each identified concern: create a `TaskCreate` sub-task, work through it with `file:line` evidence, mark done. No findings without proof.
>
> **Illustrative concern examples by category type** (not exhaustive — trust your knowledge beyond this):
>
> - _Server-side logic:_ handler/service structure conventions, validation layer placement, side-effect isolation, cross-service boundary enforcement, data-access layer separation, error propagation strategy
> - _Client-side logic:_ component lifecycle management, resource cleanup (subscriptions, listeners, timers), state management patterns, API integration layer separation, reactive stream composition
> - _Data/Schema:_ migration reversibility (rollback script), lock impact on table volume, backfill idempotency, index coverage for query patterns, deployment ordering
> - _Configuration:_ present in ALL environments? No secrets in diff? App fails fast if config missing (not silently null)? Documented in setup guide?
> - _Infrastructure:_ dev/prod parity? No hardcoded dev values (localhost, debug flags)? Pinned image/dependency versions? CI/CD secret requirements documented?
> - _Styles/Assets:_ follows project naming conventions? Uses design variables/tokens (no hardcoded magic values)? Correct scope (no global side effects from component styles)?
> - _Documentation:_ accurate? Links valid? Examples still match current code/behavior? Covers new scenarios?
> - _Tests:_ assertions verify specific outcomes (not just "no exception")? Idempotent (repeatable N times)? Covers edge cases, not just happy path?
> - _Security artifacts:_ all code paths reach the gate? Negative tests exist (unauthorized denied)? Both enforcement AND display control updated?
> - _Build/Tooling:_ rule changes apply consistently? No exceptions that silently swallow violations? Impact on CI runtime documented?

<!-- /SYNC:category-review-thinking -->

<!-- SYNC:subagent-return-contract -->

> **Sub-Agent Return Contract** — When this skill spawns a sub-agent, the sub-agent MUST return ONLY the structured envelope below. Main agent reads the envelope first, then opens the referenced report for synthesis, acceptance, deduplication, or repair planning; a full report is never pasted inline.
>
> ```markdown
> ## Sub-Agent Result: [skill-name]
>
> Status: ✅ PASS | ⚠️ PARTIAL | ❌ FAIL
> Confidence: [0-100]%
> Run ID: [stable run identifier]
> Task ID: [parent task or phase identifier]
> Attempt ID: [monotonic attempt/revision identifier]
> Target: [exact files/paths or scope] @ [target fingerprint/commit]
> Changed paths: [none | exact paths]
> Finding totals: Critical=[n] | High=[n] | Medium=[n] | Low=[n]
> Acceptance: PENDING | ACCEPTED | REJECTED — parent records the decision
>
> ### Findings (Critical/High surfaced — max 10 bullets)
>
> - [severity] [file:line] [finding]
>
> ### Gaps / Unverified
>
> - [missing host, runtime, coverage, or evidence limitation]
>
> ### Actions Taken
>
> - [file changed] [what changed]
>
> ### Blockers (if any)
>
> - [blocker description, or `none`]
>
> Full report: plans/reports/[skill-name]-[date]-[slug].md
> ```
>
> The ten-bullet limit is a transport limit, not a visibility limit: the full report may contain more than ten Medium/Low findings when no named blocker exists, and the parent MUST read it when synthesizing or deduplicating. The parent MUST reject a stale, duplicate, or superseded `Attempt ID` and MUST accept the current attempt before advancing a dependent step. Read-only leaves write repair proposals/reports only; they do not edit source, generated carriers, or user files.
>
> **Context budget** — the return payload is a SUMMARY, not a transcript: no raw file contents / full diffs / verbatim logs inline, no re-pasted source. Everything beyond the envelope lives in the incrementally-written report. A sub-agent that would exceed the summary shape MUST persist the detail and return only the pointer; bounded transport must never become bounded visibility.

<!-- /SYNC:subagent-return-contract -->

<!-- SYNC:nested-task-creation -->

> **Nested Task Expansion Contract** — For workflow-step invocation, the `[Workflow] ...` row is only a parent container; the child skill still creates visible phase tasks.
>
> 1. Call `TaskList` first. If a matching active parent workflow row exists, set `nested=true` and record `parentTaskId`; otherwise run standalone.
> 2. Create one task per declared phase before phase work. When nested, prefix subjects `[N.M] /skill-name — phase`.
> 3. When nested, link the parent with `TaskUpdate(parentTaskId, addBlockedBy: [childIds])`.
> 4. Orchestrators must pre-expand a child skill's phase list and link the workflow row before invoking that child skill or sub-agent.
> 5. Mark exactly one child `in_progress` before work and `completed` immediately after evidence is written.
> 6. Complete the parent only after all child tasks are completed or explicitly cancelled with reason.
>
> **Blocked until:** `TaskList` done, child phases created, parent linked when nested, first child marked `in_progress`.

<!-- /SYNC:nested-task-creation -->

<!-- SYNC:project-reference-docs-guide -->

> **Project Reference Docs Gate (static JIT)** — Run after task-tracking bootstrap and immediately before target/source file reads, grep, edits, tests, or analysis. Project docs override generic framework assumptions; hooks may remind or accelerate this gate, but never prove that it ran.
>
> 1. Identify scope: file types, domain area, and operation.
> 2. **Read `docs/project-config.json` first — the project's machine-readable map.** It is the single source of truth for THIS repo (modules/paths, framework + search keywords, test/E2E/integration run-commands, design system, architecture rules, workflow patterns); ground exact paths, run-commands, and conventions on it **before investigating, planning, or coding** — never assume framework defaults (`CLAUDE.md` + reference docs are derived from it). If it — or the docs index, `lessons.md`, `CLAUDE.md`, `AGENTS.md`, or any required reference doc — is missing or stale, auto-run `/project-init` or the narrow route (`/project-config`, `/docs-init`, `/scan-all`, `/scan --target=<key>`, `/claude-md-init`) first; if Codex mirrors or `AGENTS.md` are stale, ask the user to run `/sync-codex` (never auto-run it).
> 3. Required docs by trigger: always `docs/project-reference/lessons.md`; doc lookup `docs-index-reference.md`; review `code-review-rules.md`; backend/CQRS/API `backend-patterns-reference.md`; domain/entity `domain-entities-reference.md`; frontend/UI `frontend-patterns-reference.md`; styles/design `scss-styling-guide.md` + `design-system/design-system-canonical.md`; integration tests `integration-test-reference.md`; E2E `e2e-test-reference.md`; feature docs/specs `feature-spec-reference.md` + `spec-system-reference.md` + `spec-principles.md`; behavior/public-contract/spec-test-code sync `workflow-spec-test-code-cycle-reference.md`; derived spec index/ERD/reimplementation guides `spec-system-reference.md` + source Feature Specs under `docs/specs/`; architecture/new area `project-structure-reference.md`.
> 4. Read every required doc, then before target work state: `Reference docs read: ... | Not applicable: ...`. After compaction, resume, delegation, or a material context change, repeat the route and restate the set; prior conversation and hook output are not proof of current loading.
>
> **Ready when:** scope evaluated, `docs/project-config.json` consulted, required docs checked/read or setup route completed, `lessons.md` confirmed, citation emitted.

<!-- /SYNC:project-reference-docs-guide -->

<!-- SYNC:task-tracking-external-report -->

> **Task Tracking & External Report Persistence** — Bootstrap this before execution; then run project-reference doc prefetch before target/source work.
>
> 1. Create a small task breakdown before target file reads, grep, edits, or analysis. On context loss, inspect the current task list first.
> 2. Mark one task `in_progress` before work and `completed` immediately after evidence; never batch transitions.
> 3. For plan/review work, create `plans/reports/{skill}-{YYMMDD}-{HHmm}-{slug}.md` before first finding.
> 4. Append findings after each file/section/decision and synthesize from the report file at the end.
> 5. Final output cites `Full report: plans/reports/{filename}`.
>
> **Blocked until:** task breakdown exists, report path declared for plan/review work, first finding persisted before the next finding.

<!-- /SYNC:task-tracking-external-report -->

<!-- SYNC:critical-thinking-mindset -->

> **Critical Thinking Mindset** — Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence >80% to act.
> **Anti-hallucination:** Never present guess as fact — cite sources for every claim, admit uncertainty freely, self-check output for errors, cross-reference independently, stay skeptical of own confidence — certainty without evidence root of all hallucination.

<!-- /SYNC:critical-thinking-mindset -->

<!-- SYNC:sequential-thinking-protocol -->

> **Sequential Thinking Protocol** — Structured multi-step reasoning for complex/ambiguous work. Use when planning, reviewing, debugging, or refining ideas where one-shot reasoning is unsafe.
>
> **Trigger when:** complex problem decomposition · adaptive plans needing revision · analysis with course correction · unclear/emerging scope · multi-step solutions · hypothesis-driven debugging · cross-cutting trade-off evaluation.
>
> **Format (explicit mode — visible thought trail):**
>
> 1. `Thought N/M: [aspect]` — one aspect per thought, state assumptions/uncertainty
> 2. `Thought N/M [REVISION of Thought K]: ...` — when prior reasoning invalidated; state Original / Why revised / Impact
> 3. `Thought N/M [BRANCH A from Thought K]: ...` — explore alternative; converge with decision rationale
> 4. `Thought N/M [HYPOTHESIS]: ...` then `[VERIFICATION]: ...` — test before acting
> 5. `Thought N/N [FINAL]` — only when verified, all critical aspects addressed, confidence >80%
>
> **Mandatory closers:** Confidence % stated · Assumptions listed · Open questions surfaced · Next action concrete.
>
> **Stop conditions:** confidence <80% on any critical decision → escalate via AskUserQuestion · ≥3 revisions on same thought → re-frame the problem · branch count >3 → split into sub-task.
>
> **Implicit mode:** apply methodology internally without visible markers when adding markers would clutter the response (routine work where reasoning aids accuracy).
>
> **Deep-dive:** see `/sequential-thinking` skill (`.claude/skills/sequential-thinking/SKILL.md`) for worked examples (API design, debugging, architecture), advanced techniques (spiral refinement, hypothesis testing, convergence), and meta-strategies (uncertainty handling, revision cascades).

<!-- /SYNC:sequential-thinking-protocol -->

<!-- SYNC:evidence-based-reasoning -->

> **Evidence-Based Reasoning** — Speculation is FORBIDDEN. Every claim needs proof.
>
> 1. Cite `file:line`, grep results, or framework docs for EVERY claim
> 2. Declare confidence: >80% act freely, 60-80% verify first, <60% DO NOT recommend
> 3. Cross-service validation required for architectural changes
> 4. "I don't have enough evidence" is valid and expected output
>
> **BLOCKED until:** `- [ ]` Evidence file path (`file:line`) `- [ ]` Grep search performed `- [ ]` 3+ similar patterns found `- [ ]` Confidence level stated
>
> **Forbidden without proof:** "obviously", "I think", "should be", "probably", "this is because"
> **If incomplete →** output: `"Insufficient evidence. Verified: [...]. Not verified: [...]."`

<!-- /SYNC:evidence-based-reasoning -->

<!-- SYNC:design-patterns-quality -->

> **Design Patterns Quality** — Priority checks for every code change:
>
> 1. **DRY via OOP:** Identify classes/modules with the same purpose, naming pattern, or lifecycle. Apply your knowledge of the project's language/framework to determine the idiomatic abstraction (base class, mixin, trait, protocol, decorator). 3+ similar patterns → extract to shared abstraction.
> 2. **Right Responsibility:** Logic in LOWEST layer (Entity > Domain Service > Application Service > Controller). Never business logic in controllers.
> 3. **SOLID:** Single responsibility (one reason to change). Open-closed (extend, don't modify). Liskov (subtypes substitutable). Interface segregation (small interfaces). Dependency inversion (depend on abstractions).
> 4. **After extraction/move/rename:** Grep ENTIRE scope for dangling references. Zero tolerance.
> 5. **YAGNI gate:** Recommend extraction when 3+ similar patterns exist OR an evidenced consumer boundary/substitution need justifies it; do not create patterns for hypothetical future use.
> 6. **Purpose-oriented naming protocol:** Name public or cross-layer abstractions by the capability, domain purpose, or contract consumers rely on—not the current provider, SDK, framework, database, or transport. `IStorage`/`Storage` → `AzureBlobStorage`; use `IAzureStorage` only when Azure-specific semantics are intentionally part of the contract. — why: provider-coupled names make an implementation replacement look like a contract change.
> 7. **Contract-fit check:** Read callers and every implementation before judging a name; narrow an over-broad abstraction (`IObjectStore`, `DocumentStore`) instead of rewarding a generic name that lies about behavior. — why: a name cannot be validated from the declaration alone.
> 8. **Mechanism/generic-name smell:** Treat `Manager`, `Helper`, `Utils`, `Data`, `Thing`, `Service`, `Interface`, type decorations, and unexplained abbreviations as review signals—not automatic defects; flag them only when they hide purpose, scope, or responsibility. — why: blanket word bans replace judgment with another naming convention.
> 9. **Concrete implementation names:** Provider, strategy, transport, or test-double names are valid on concrete types when they distinguish real behavior (`AzureBlobStorage`, `InMemoryStorage`, `RetryingStorage`); keep those details out of the caller-facing contract unless the contract promises them. — why: implementation names should explain the selected behavior while callers depend on stable semantics.
> 10. **Language convention:** Preserve local interface syntax and naming style; `.NET` `I` prefixes and Google TypeScript's unmarked interfaces are both valid local conventions. — why: purpose-oriented naming is universal, marker syntax is ecosystem-specific.
>
> **Anti-patterns to flag:** God Object, Copy-Paste inheritance, Circular Dependency, Leaky Abstraction.
>
> **Serial Attention for Design Quality** — Scan one quality dimension at a time (serial passes), not all concerns at once. — why: split attention misses violations that single-focus passes catch.
>
> 1. **Identify applicable dimensions** — Based on the code's language, domain, and patterns, determine which quality dimensions apply: DRY, SOLID principles (SRP/OCP/LSP/ISP/DIP), OOP idioms, cohesion/coupling, GRASP, Law of Demeter, CQRS invariants, etc. Your list is NOT fixed — derive from what the code actually does.
> 2. **One focused pass per dimension** — Dedicate single-focus attention to EACH dimension in sequence. Do NOT mix concerns across passes.
> 3. **Threshold: 3+ similar patterns = MANDATORY extraction** — Not optional suggestion. Flag as mandatory structural fix requiring action.
> 4. **2+ violations of same kind = structural finding** — Report as "pattern problem" needing architectural resolution, not a list of individual instances.

<!-- /SYNC:design-patterns-quality -->

<!-- SYNC:complexity-prevention -->

> **Complexity Prevention (Ousterhout)** — MANDATORY. Measure code by cost of change: one business change should map to one code change. Flag ALL of the following in review:
>
> 1. **Change amplification** — small business change forces edits in >3 places → structural flaw. Count edit sites for a plausible future change (add variant, add field, add authorization). >3 = reject.
> 2. **Cognitive load** — reader must hold too much context to safely modify. Flag deep inheritance, long parameter lists, boolean traps, implicit ordering dependencies.
> 3. **Cross-cutting duplication at entry points** — logging, error handling, validation, auth, transactions reimplemented per controller/handler/route. Lift to middleware / interceptor / filter / decorator / aspect.
> 4. **Leaked implementation technology** — repos returning `IQueryable`/`QuerySet`/`Criteria`/raw cursors/ORM entities to callers. Return finished results + intent-revealing methods (`GetActiveVipUsers()` not `Query()`).
> 5. **Type-switch scattering** — `switch`/`if`-chains on enum/discriminator in >1 place. New variant = new file, not N edits. One factory/registry switch at the boundary OK; scattered switches = reject.
> 6. **Anemic models** — domain objects with only getters/setters, logic floats in services. Move invariants/behavior onto the object (`order.Checkout()`, not `order.Status = ...`).
> 7. **Primitive obsession** — raw `string`/`int`/`decimal` for account numbers, emails, money, percentages, date ranges, with re-validation at every entry. Wrap in value objects / records / structs that validate once at construction.
> 8. **Inline cross-cutting concerns** — authorization/tenant isolation/audit/sanitization hand-written at top of every handler. Flag intent with declarative markers (`@RequirePermission("Order.Delete")`), enforce once centrally.
> 9. **Shallow modules** — tiny class, big interface (many public methods, many flags, many ctor params) wrapping little logic. A module is deep when a small interface hides a lot of implementation. If interface ≈ implementation cost to learn → inline.
> 10. **Missing base class for repeated component/handler lifecycle** — 3+ forms/CRUD handlers/list views reimplementing loading/dirty/submit/pagination → extract to base class / hook / composable / mixin / trait.
> 11. **Premature vs delayed abstraction** — rule-of-three. First occurrence: write it. Second: notice duplication. Third: extract. Don't build generic frameworks before real variation; don't copy-paste for the 4th time.
> 12. **Embedded utility logic not extracted to helpers** — inline paging loops (`while (hasMore) { skip += take; ... }`), ad-hoc datetime math, string parsing/formatting, collection partitioning, retry/backoff loops, URL/query-string building. If the algorithm is non-trivial AND stack-generic (not business-specific), extract to `util`/`helper`/`extensions` and let consumers call one line. Inline duplicates → duplicated bug surface.
> 13. **Logic in wrong (higher) layer — downshift to callee** — business/derivation logic written in the caller when the callee owns the data. Defaults: Controller code that should be App Service. App Service code that should be Domain Service or Entity. Component code that should be ViewModel/Store/Service. Caller reaching into callee's data shape to compute something → move the computation behind an intent-revealing method on the callee. Lowest responsible layer wins (Entity > Domain Service > App Service > Controller · Model/VM > Store > Component). Higher-layer placement = duplicated logic when a sibling caller needs the same thing.
> 14. **Owner owns the rule — extract on first write** — if a caller inlines logic that derives, normalizes, validates, or computes from another type's data, MOVE it to the owning type. Single use is sufficient — the trigger is wrong responsibility, not duplication. Sibling callers always arrive; inline copies drift silently with no compile error and no name to grep. **Common offenders:** _Backend_ — inlined rules in application-layer handlers / commands / queries / services / controllers that belong on the domain entity / value object / domain service. _Frontend_ — inlined derivations / formatting / validation in components that belong on the model / store / view-model / API service. **Fix:** name the rule once as a method (static or instance) on the owning type; callers invoke by name. Future variant → SECOND named method on the owner, never an inline near-duplicate. **Right responsibility first; reuse is the consequence.**
>
> **Extraction target — where the named rule lives:**
>
> | Shape of the rule                             | Goes to                       |
> | --------------------------------------------- | ----------------------------- |
> | Pure function over an entity's own data       | static method on the entity   |
> | Behavior that mutates / guards entity state   | instance method on the entity |
> | Always-true invariant on a primitive value    | value object constructor      |
> | Needs DI (repo / settings / clock)            | helper class registered in DI |
> | Domain-agnostic algorithm reused across types | util / extension method       |
> | Pure shape / projection conversion            | DTO mapping                   |
>
> **Pre-commit edit-site test (reject if answer is "many"):**
>
> | Change Scenario                                 | Should touch              |
> | ----------------------------------------------- | ------------------------- |
> | Add new variant (customer type, payment method) | 1 new file                |
> | Change HTTP error response format               | 1 middleware/filter       |
> | Add timestamp field to every persisted entity   | 1 base entity/interceptor |
> | Add authorization to a new endpoint             | 1 declarative marker      |
> | Swap database/ORM                               | Data layer only           |
> | Change business calculation rule                | 1 method on owning entity |
> | Add loading indicator pattern to forms          | 1 base component/hook     |
> | Add validation rule to a domain primitive       | 1 value-object ctor       |
> | Change paging/retry/datetime algorithm          | 1 helper/util function    |
> | Change a derivation of entity data              | 1 method on the entity    |
>
> **Operating heuristics:**
>
> - Write the call site first.
> - Count edit sites for plausible future change.
> - Prefer removing code over adding it.
> - Surface assumptions at boundaries, hide details inside.
> - **Pre-reuse scan** — before writing a non-trivial block, grep for similar algorithms (`while.*skip`, `DateTime.*Add`, `split`/`join` chains, paging loops, retry loops). Match existing helper → call it. None exists but pattern is stack-generic → extract to util before second caller appears.
> - **Layer placement test** — ask "if a sibling caller needed this tomorrow, would they re-derive it?" If yes, the logic is in the wrong layer. Move it down.
> - **Open-case-for-future-reuse** — if reviewer spots a block that is likely to appear in another feature (domain-agnostic algorithm, shared lifecycle, recurring derivation), do NOT rationalize with pure YAGNI. Either extract now (if cheap) or create a tracked TODO with the exact extraction target so the second caller does not duplicate silently. Silent duplication is the default failure mode.
> - When in doubt ask: "What would need to change if the requirement shifts?"
>
> **The measure of good code is the cost of change.** Not shortest. Not cleverest. Not most abstracted. Cheapest to safely modify having read a small local portion.

<!-- /SYNC:complexity-prevention -->

<!-- SYNC:double-round-trip-review -->

> **Validated-Finding Fix + Full Re-Review Loop** — Re-review is triggered by a validated finding fix cycle or an explicitly declared independent-pass minimum, not by a round number alone. Review purpose: `review → validate findings → fix validated findings that block the current round → full re-review` until a complete review pass clears the round's exit bar (see **Severity floor** below). **A clean review ENDS the loop once the persisted `minRounds` is met (default 1); an explicitly declared minimum such as 2 still requires that independent pass.**
>
> _aka **Self-Review Convergence Loop**._ The name is historical — there is **NO 2-round cap**; "double-round-trip" only means a validated-finding fix cycle forces at least one fresh re-review. It runs until the current round's exit bar is clear (round 1: zero findings; round 2+: zero CRITICAL/HIGH/MEDIUM, LOW deferred), bounded by the **3-round ceiling** below.
>
> **Round cap — 3 rounds MAX (a ceiling, NEVER a target).** A clean pass ENDS the loop at ANY round once `round >= minRounds` — round 1 included with the default minimum; the cap never obliges extra rounds. Hitting round 3 with blocking findings still open (severity floor applied) → **STOP and escalate via `AskUserQuestion`** with the still-open findings listed; NEVER emit a silent "good enough" PASS on cap exhaustion, and NEVER let the cap substitute for the clean-review requirement. The 2-repeated-no-progress blocker rule stays an EARLIER exit — escalate at whichever trips first.
>
> **Severity floor — from round 2, LOW stops blocking.** The exit bar tightens after the first review pass, so the loop converges on consequence instead of spinning on polish:

> Define one predicate everywhere: `blocking_findings(round, findings)` returns all validated findings in round 1 and only validated CRITICAL/HIGH/MEDIUM findings in rounds 2–3. A binary gate (test-green, security must-fix, required artifact) is exempt only when its owning invariant explicitly says so; in practice binary gates always remain blocking when they fail.
>
> | Round | Exit bar — loop ENDS when the fresh full review has… | Must be fixed to continue |
> | --- | --- | --- |
> | 1 | zero validated findings at ANY severity | CRITICAL · HIGH · MEDIUM · LOW |
> | 2–3 | zero validated CRITICAL / HIGH / MEDIUM findings — **LOW-only clears the severity bar** | CRITICAL · HIGH · MEDIUM only |
>
> From round 2 onward LOW findings are **NOT required to be fixed**: a round whose validated findings are ALL LOW **ENDS the loop once the persisted minimum is met** — do not open another fix/re-review round for them. Severity tiers are `SYNC:severity-rubric` (CRITICAL block-merge · HIGH must-fix · MEDIUM must clear the current round · LOW record/defer); round 1 remains strict, so a LOW found initially is still validated and fixed when warranted before the floor can apply.
>
> **Severity-floor rules:**
>
> - **Never silently drop a deferred LOW.** Every unfixed LOW is listed in the final report under `## Deferred LOW Findings (severity floor, round ≥2)` with file, line, and description, so the owner can schedule it. Dropping it from the report is a protocol violation, not a clean pass.
> - **Never re-tier a finding to trigger the exit.** Downgrading a real CRITICAL/HIGH/MEDIUM to LOW so the loop can end is a FALSE PASS. Severity is set by consequence per `SYNC:severity-rubric` before the round bar is applied — never after, and never with the exit in view. — why: a floor that can be reached by relabeling is not a floor.
> - **The floor bounds the loop, not the standard.** It ends *iteration*; it never authorizes shipping a known CRITICAL/HIGH/MEDIUM, and it never lowers the finding-survival bar that admits a finding in the first place.
> - **The floor never applies to a hard gate.** Test-green gates (a suite must actually pass), security must-fix gates, and any gate whose criterion is binary rather than severity-rated are unaffected — a failing test is a failure, not a LOW finding.
>
> **Universal scope (any new output/judgment):** any newly produced output or judgment gets **≥1 self-review**; any **new judgment** gets **≥1 `/why-review --validate-findings` pass**; anything flagged to re-check is re-checked **≥1 time** — before that output is treated as final. This loop is the default convergence contract for ANY work-producing skill, not review skills only.
>
> **Routing invariant (author-facing):** a skill that validates findings MUST route them through `/why-review --validate-findings` (the terminal validator) — NEVER fork an inline finding-validation. Routing through why-review is what makes the finding-survival bar and this loop apply; the `verify-review-validate-coverage` sensor enforces this exact route mechanically.
>
> **Round 1:** Main-session review. Read target files, build understanding, note issues. Output findings + verdict (PASS / FAIL).
>
> **Decision after Round 1:**
>
> - **No issues found (PASS, zero findings)** → review ENDS if `round >= minRounds`; otherwise perform the explicitly required independent pass. Do NOT invent a confirmation pass.
> - **`blocking_findings(round, findings)` is non-empty** → run the active review skill's findings-validation gate first; for review skills the default gate is `/why-review --validate-findings <report-path>`. Fix only validated findings that block the current round, then restart the full review protocol from the beginning with a fresh task breakdown.
>
> **Fresh full re-review after every fix cycle:** Re-run the whole review protocol over the current full target. When sub-agents are part of that protocol, spawn NEW `Agent` calls — never reuse prior agents. Reviewers re-read ALL files from scratch with ZERO memory of prior rounds. See `SYNC:fresh-context-review` for the spawn mechanism and `SYNC:review-protocol-injection` for the canonical Agent prompt template. Each fresh full review must catch:
>
> - Cross-cutting concerns missed in the prior round
> - Interaction bugs between changed files
> - Convention drift (new code vs existing patterns)
> - Missing pieces that should exist but don't
> - Subtle edge cases the prior round rationalized away
> - Regressions introduced by the fixes themselves
>
> **Loop termination:** After each full re-review, repeat the same decision against **that round's exit bar**: bar cleared and persisted minimum met → END; blocking findings remain → validate findings → fix → restart from the first review phase. Round 1 clears only on zero findings at any severity; **from round 2 the bar is zero CRITICAL/HIGH/MEDIUM, so a LOW-only round ENDS the loop once the persisted minimum is met** (deferred LOWs go in the report). Capped at **3 rounds**. Escalate via `AskUserQuestion` at whichever comes first: the same validated finding repeats for 2 full invocations with no progress · a fix requires product/owner input · round 3 completes with CRITICAL/HIGH/MEDIUM still open. NEVER loop past 3 rounds, and NEVER convert cap exhaustion into a PASS.
>
> **Rules:**
>
> - A clean Round 1 ENDS the review when `minRounds=1`; an explicitly declared `minRounds=2` requires the independent second pass
> - From round 2 on, a round whose validated findings are ALL LOW ENDS the loop once the persisted minimum is met — never open round N+1 to fix LOW alone; list those LOWs as deferred instead
> - NEVER re-tier a CRITICAL/HIGH/MEDIUM down to LOW to reach the round-2 exit — severity is assigned by consequence before the bar is applied
> - NEVER fix unvalidated findings; validate first using the caller's validation gate
> - Every surviving finding must additionally clear the **finding-survival bar** defined in why-review's Findings Validation Routine (a deliberately higher bar than the generic act-gate — "keep this finding?" is a stricter question than "act on this evidence?"); a finding below the bar is demoted or dropped, not kept
> - NEVER skip the full re-review after a fix cycle (every fix invalidates the prior verdict)
> - NEVER reuse a sub-agent across rounds — every iteration that uses sub-agents spawns NEW Agent calls
> - Main agent READS sub-agent reports but MUST NOT filter, reinterpret, or override findings
> - The 3-round cap NEVER replaces the clean-review requirement — it bounds runaway looping, it does not authorize shipping an un-clean review; a clean pass ends the loop early once the persisted minimum is met, and cap exhaustion escalates rather than passes
> - Enforce the round cap of 3 alongside the 2 repeated-no-progress blocker rule; both are escalation triggers, neither is a completion criterion
> - Persist completed rounds, repeated blockers, findings and the explicit minimum in the owning run's `review-policy.cjs` record. Resume that record after interruption; target changes invalidate evidence and acceptance but preserve the bounded round budget. In-flight attempt IDs may be session-local; they do not replace or reset completed-round state
> - Final verdict must incorporate ALL rounds executed
>
> **Report must include `## Round N Findings (Fresh Sub-Agent)` for every round N≥2 that was executed, plus `## Deferred LOW Findings (severity floor, round ≥2)` whenever the loop ended on the round-2+ bar with LOWs still open.**

<!-- /SYNC:double-round-trip-review -->


<!-- SYNC:fresh-context-review -->

> **Fresh Context Re-Review** — Eliminate orchestrator confirmation bias after fixes by restarting the full review with isolated sub-agents where applicable. A report-only/read-only reviewer never edits source, generated output, or user data: it validates and records the finding/repair handoff, then returns to the caller, which owns the fix and any re-review.
>
> **Why:** The main agent knows what it (or `/feature-implement`) just fixed and rationalizes findings accordingly. A fresh sub-agent has ZERO memory, re-reads from scratch, and catches what the main agent dismissed. Sub-agent bias is mitigated by (1) fresh context, (2) verbatim protocol injection, (3) main agent not filtering the report.
>
> **When:** After a validated-finding fix cycle, or to satisfy an explicitly declared independent-pass `minRounds`. A review round that finds zero issues ENDS the loop once that persisted minimum is met — do NOT invent a confirmation sub-agent. A review round that finds issues triggers: validate findings → fix → full review restart from the first phase.
>
> **How:**
>
> 1. Start a NEW full review invocation/task breakdown; when that protocol calls for agents, spawn NEW `Agent` tool calls — use `code-reviewer` subagent_type for code reviews, `general-purpose` for plan/doc/artifact reviews
> 2. Inject ALL required review protocols VERBATIM into the prompt — see `SYNC:review-protocol-injection` for the full list and template. Never reference protocols by file path; AI compliance drops behind file-read indirection (see `SYNC:shared-protocol-duplication-policy`)
> 3. Sub-agent re-reads ALL target files from scratch via its own tool calls — never pass file contents inline in the prompt
> 4. Sub-agent writes structured report to `plans/reports/{review-type}-round{N}-{date}.md`
> 5. Main agent reads the report, integrates findings into its own report, DOES NOT override or filter
>
> **Rules:**
>
> - SKIP fresh sub-agent when the prior full review found zero issues AND the persisted `minRounds` is met (no fixes or required independent pass = nothing new to verify)
> - NEVER skip the full review restart after a fix cycle — every fix invalidates the prior verdict
> - NEVER reuse a sub-agent across rounds — every fresh round spawns a NEW `Agent` call
> - Continue until a complete full review pass clears that round's exit bar per `SYNC:double-round-trip-review`: **round 1** → zero findings at any severity; **round 2+** → zero CRITICAL/HIGH/MEDIUM, so a round whose validated findings are ALL LOW ENDS the loop once the persisted minimum is met (list those LOWs as deferred instead of spawning another round). If the same validated blocker repeats across 2 full invocations with no progress, escalate via `AskUserQuestion`. **Read-only/report-only role boundary:** when this block is carried by a security auditor or another report-only role, “fix” means return the validated repair proposal to the parent; do not modify source, generated carriers, or user data and do not restart the review locally.
> - Persist completed rounds, repeated blockers, findings and the explicit minimum in the owning run's `review-policy.cjs` record. Resume that record after interruption; target changes invalidate evidence and acceptance but preserve the bounded round budget. In-flight attempt IDs may be session-local; they do not replace or reset completed-round state

<!-- /SYNC:fresh-context-review -->

<!-- SYNC:review-protocol-injection -->

> **Review Protocol Injection** — Every fresh sub-agent review prompt MUST embed 11 protocol blocks VERBATIM. The template below has ALL 11 bodies already expanded inline. Copy the template wholesale into the Agent call's `prompt` field at runtime, replacing only the `{placeholders}` in Task / Round / Reference Docs / Target Files / Output sections with context-specific values. Do NOT touch the embedded protocol sections.
>
> **Why inline expansion:** Placeholder markers would force file-read indirection at runtime. AI compliance drops significantly behind indirection (see `SYNC:shared-protocol-duplication-policy`). Therefore the template carries all 11 protocol bodies pre-embedded.

### Subagent Type Selection

- `code-reviewer` — for code reviews (reviewing source files, git diffs, implementation)
- `general-purpose` — for plan / doc / artifact reviews (reviewing markdown plans, docs, specs)

### Canonical Agent Call Template (Copy Verbatim)

```
Agent({
  description: "Fresh Round {N} review",
  subagent_type: "code-reviewer",
  prompt: `
## Task
{review-specific task — e.g., "Review all uncommitted changes for code quality" | "Review plan files under {plan-dir}" | "Review integration tests in {path}"}

## Round
Round {N}. You have ZERO memory of prior rounds. Re-read all target files from scratch via your own tool calls. Do NOT trust anything from the main agent beyond this prompt.

## Protocols (follow VERBATIM — these are non-negotiable)

### Spec ↔ Tests ↔ Code Triangulation
DO THIS FIRST — before any per-protocol check below. The review target is the WHOLE PACKAGE, not the diff alone: load the behavior's spec (§3 ACs / §4 BRs / §8 TCs), its tests, and the changed code TOGETHER, and reason about their mutual consistency BEFORE judging any one in isolation.
1. Locate all three faces: the Feature Spec section(s) governing the changed behavior, the tests that guard it, and the production code that implements it. A missing face is itself a finding (SPEC-GAP / TEST-GAP / DEAD-SPEC).
2. Triangulate pairwise — every disagreement is a finding; classify which face is wrong:
   - code vs spec: behavior the code does that no §3/§4/§8 rule describes → CODE-EXTRA or SPEC-STALE; a [HARD] §4 rule or §5 invariant with no enforcing code path → CODE-WRONG.
   - tests vs spec: a §8 TC with no test, or a test asserting behavior no TC/rule names → TEST-GAP or SPEC-SILENT.
   - tests vs code: a changed code path with no covering test → TEST-GAP; a test that still passes against a deliberately broken invariant → WEAK-TEST (apply the mutation thinking in Bug Detection).
3. Hidden-rule capture: any invariant the code enforces but the spec never states (SPEC-SILENT) MUST be surfaced as a finding to add into §3/§4/§8 AND guarded with a test — the enrichment loop, never a silent pass.
4. Only after the three faces agree — or every disagreement is logged as a finding — proceed to the per-protocol checks below; when enrichment adds spec/test content, re-review the package against the enriched spec.
NEVER mark review PASS while any spec/test/code face disagrees without a logged finding. The diff is the entry point; the package is the unit of judgment.

### Evidence-Based Reasoning
Speculation is FORBIDDEN. Every claim needs proof.
1. Cite file:line, grep results, or framework docs for EVERY claim
2. Declare confidence: >80% act freely, 60-80% verify first, <60% DO NOT recommend
3. Cross-service validation required for architectural changes
4. "I don't have enough evidence" is valid and expected output
BLOCKED until: Evidence file path (file:line) provided; Grep search performed; 3+ similar patterns found; Confidence level stated.
Forbidden without proof: "obviously", "I think", "should be", "probably", "this is because".
If incomplete → output: "Insufficient evidence. Verified: [...]. Not verified: [...]."

### Bug Detection
MUST check categories 1-4 for EVERY review. Never skip.
1. Null Safety: Can params/returns be null? Are they guarded? Optional chaining gaps? .find() returns checked?
2. Boundary Conditions: Off-by-one (< vs <=)? Empty collections handled? Zero/negative values? Max limits?
3. Error Handling: Try-catch scope correct? Silent swallowed exceptions? Error types specific? Cleanup in finally?
4. Resource Management: Connections/streams closed? Subscriptions unsubscribed on destroy? Timers cleared? Memory bounded?
5. Concurrency (if async): Missing await? Race conditions on shared state? Stale closures? Retry storms?
6. Stack-Specific: Check the configured language/runtime pitfalls and framework-specific failure modes discovered from local code.
Classify every finding by consequence using `SYNC:severity-rubric` (never by effort): CRITICAL = immediate material security/safety/data-loss risk or failed binary gate → block; HIGH = material correctness, contract, privacy, or authority risk → must fix; MEDIUM = bounded consequential edge/resilience/maintainability gap → must clear the current round, or escalate with an explicit residual-risk follow-up that does not create a clean pass; LOW = non-blocking polish with no credible present impact → record/defer from round 2; `NOT VERIFIABLE` is unresolved evidence, not LOW.

### Design Patterns Quality
Priority checks for every code change:
1. DRY via OOP: Same-suffix classes (*Entity, *Dto, *Service) MUST share base class. 3+ similar patterns → extract to shared abstraction.
2. Right Responsibility: Logic in LOWEST layer (Entity > Domain Service > Application Service > Controller). Never business logic in controllers.
3. SOLID: Single responsibility (one reason to change). Open-closed (extend, don't modify). Liskov (subtypes substitutable). Interface segregation (small interfaces). Dependency inversion (depend on abstractions).
4. After extraction/move/rename: Grep ENTIRE scope for dangling references. Zero tolerance.
5. YAGNI gate: Recommend extraction when 3+ similar patterns exist OR an evidenced consumer boundary/substitution need justifies it; do not create patterns for hypothetical future use.
6. Purpose-oriented naming: Name public or cross-layer abstractions by the capability, domain purpose, or contract consumers rely on—not the current provider, SDK, framework, database, or transport. `IStorage`/`Storage` → `AzureBlobStorage`; use `IAzureStorage` only when Azure-specific semantics are intentionally part of the contract.
7. Contract-fit check: Read callers and every implementation before judging a name; narrow an over-broad abstraction (`IObjectStore`, `DocumentStore`) instead of rewarding a generic name that lies about behavior.
8. Mechanism/generic-name smell: Treat `Manager`, `Helper`, `Utils`, `Data`, `Thing`, `Service`, `Interface`, type decorations, and unexplained abbreviations as review signals—not automatic defects; flag them only when they hide purpose, scope, or responsibility.
9. Concrete implementation names: Provider, strategy, transport, or test-double names are valid on concrete types when they distinguish real behavior (`AzureBlobStorage`, `InMemoryStorage`, `RetryingStorage`); keep those details out of the caller-facing contract unless the contract promises them.
10. Language convention: Preserve local interface syntax and naming style; `.NET` `I` prefixes and Google TypeScript's unmarked interfaces are both valid local conventions.
Anti-patterns to flag: God Object, Copy-Paste inheritance, Circular Dependency, Leaky Abstraction.

### Logic & Intention Review
Verify WHAT code does matches WHY it was changed.
1. Change Intention Check: Every changed file MUST serve the stated purpose. Flag unrelated changes as scope creep.
2. Happy Path Trace: Walk through one complete success scenario through changed code.
3. Error Path Trace: Walk through one failure/edge case scenario through changed code.
4. Acceptance Mapping: If plan context available, map every acceptance criterion to a code change.
5. Tests Verify Intent: For test/spec changes, verify tests name the protected business rule or invariant and would fail if that intent breaks.
6. Migration Test Exclusion: Do not write tests for migration code. Schema/data migrations are one-time execution paths, not core application logic.
NEVER mark review PASS without completing both traces (happy + error path).

### Test Spec Verification
Map changed code to test specifications.
1. Identify the project's test/spec format from existing docs, test-case files, BDD feature files, or spec folders.
2. Every changed code path MUST map to a corresponding test case/spec (or flag as "needs test case").
3. New functions/endpoints/handlers → flag for test spec creation.
4. Migration files are excluded from test/spec creation; schema/data migrations are one-time execution paths, not core application logic.
5. If spec evidence fields exist, verify they point to actual code (file:line, not stale references).
6. Verify each meaningful test case names the business intent/invariant; flag behavior-only cases that only mirror implementation details.
7. Auth/data changes → verify corresponding authorization and data-state test cases exist.
8. If no specs exist for a changed path → log the gap and recommend the project's test-spec workflow.
NEVER skip test mapping. Untested code paths are the #1 source of production bugs.

### Behavioral Delta Matrix
MANDATORY for any bugfix review. Produce input-state × pre-fix × post-fix × delta table BEFORE writing verdict.
- Minimum 3 rows; include at least one row OUTSIDE the original bug report.
- Any "REGRESSION" delta → review returns FAIL until a preservation test is added.
- Narrative descriptions do NOT substitute for the matrix.
Example rows (external-record sync fix):
| Input                 | Pre-fix | Post-fix                  | Delta      |
| --------------------- | ------- | ------------------------- | ---------- |
| Record exists (valid) | Reused  | Always recreated → orphan | REGRESSION |
| Record missing (404)  | Error   | Recreated                 | Fixed      |

### Fix-Layer Accountability
NEVER fix at the crash site. Trace the full flow, fix at the owning layer. The crash site is a SYMPTOM, not the cause.
MANDATORY before ANY fix:
1. Trace full data flow — Map the complete path from data origin to crash site across ALL layers (storage → backend → API → frontend → UI). Identify where bad state ENTERS, not where it CRASHES.
2. Identify the invariant owner — Which layer's contract guarantees this value is valid? Fix at the LOWEST layer that owns the invariant, not the highest layer that consumes it.
3. One fix, maximum protection — If fix requires touching 3+ files with defensive checks, you are at the wrong layer — go lower.
4. Verify no bypass paths — Confirm all data flows through the fix point. Check for direct construction skipping factories, clone/spread without re-validation, raw data not wrapped in domain models, mutations outside the model layer.
BLOCKED until: Full data flow traced (origin → crash); Invariant owner identified with file:line evidence; All access sites audited (grep count); Fix layer justified (lowest layer that protects most consumers).
Anti-patterns (REJECT): "Fix it where it crashes" (crash site ≠ cause site, trace upstream); "Add defensive checks at every consumer" (scattered defense = wrong layer); "Both fix is safer" (pick ONE authoritative layer).

### Rationalization Prevention
AI skips steps via these evasions. Recognize and reject:
- "Too simple for a plan" → Simple + wrong assumptions = wasted time. Plan anyway.
- "I'll test after" → RED before GREEN. Write/verify test first.
- "Already searched" → Show grep evidence with file:line. No proof = no search.
- "Just do it" → Still need TaskCreate. Skip depth, never skip tracking.
- "Just a small fix" → Small fix in wrong location cascades. Verify file:line first.
- "Code is self-explanatory" → Future readers need evidence trail. Document anyway.
- "Combine steps to save time" → Combined steps dilute focus. Each step has distinct purpose.

### Graph-Assisted Investigation
MANDATORY when .code-graph/graph.db exists.
HARD-GATE: MUST run at least ONE graph command on key files before concluding any investigation.
Pattern: Grep finds files → trace --direction both reveals full system flow → Grep verifies details.
- Investigation: trace --direction both on 2-3 entry files
- Fix/Debug: callers_of on buggy function + tests_for
- Feature/Enhancement: connections on files to be modified
- Code Review: tests_for on changed functions
- Blast Radius: trace --direction downstream
CLI: python .claude/scripts/code_graph {command} --json. Use --node-mode file first (10-30x less noise), then --node-mode function for detail.

### Understand Code First
HARD-GATE: Do NOT write, plan, or fix until you READ existing code.
1. Search 3+ similar patterns (grep/glob) — cite file:line evidence.
2. Read existing files in target area — understand structure, base classes, conventions.
3. Run python .claude/scripts/code_graph trace <file> --direction both --json when .code-graph/graph.db exists.
4. Map dependencies via connections or callers_of — know what depends on your target.
5. Write investigation to .ai/workspace/analysis/ for non-trivial tasks (3+ files).
6. Re-read analysis file before implementing — never work from memory alone.
7. NEVER invent new patterns when existing ones work — match exactly or document deviation.
BLOCKED until: Read target files; Grep 3+ patterns; Graph trace (if graph.db exists); Assumptions verified with evidence.

## Reference Docs (READ before reviewing)
- `.claude/docs/development-rules.md` — canonical development rules, code-quality guidelines, and pre-commit checklist
- docs/project-reference/code-review-rules.md
- {skill-specific reference docs — e.g., integration-test-reference.md for integration-test-review; backend-patterns-reference.md for backend reviews; frontend-patterns-reference.md for frontend reviews}

## Target Files
{explicit file list OR "run git diff to see uncommitted changes" OR "read all files under {plan-dir}"}

## Output
Write a structured report to plans/reports/{review-type}-round{N}-{date}.md with sections:
- Status: PASS | FAIL
- Issue Count: {number}
- Critical Issues (with file:line evidence)
- High Priority Issues (with file:line evidence)
- Medium / Low Issues
- Cross-cutting findings

Return the report path and status to the main agent.
Every finding MUST have file:line evidence. Speculation is forbidden.
`
})
```

### Rules

- DO copy the template wholesale — including all 11 embedded protocol sections
- DO replace only the `{placeholders}` in Task / Round / Reference Docs / Target Files / Output sections with context-specific content
- DO choose `code-reviewer` subagent_type for code reviews and `general-purpose` for plan / doc / artifact reviews
- DO NOT paraphrase, summarize, or skip any protocol section
- DO NOT pass file contents inline — the sub-agent reads via its own tool calls so it has a fresh context
- DO NOT reference protocols by file path or tag name — the bodies are already embedded above
- DO NOT introduce placeholder markers for the protocols — they must stay literally expanded

<!-- /SYNC:review-protocol-injection -->

<!-- SYNC:rationalization-prevention -->

> **Rationalization Prevention** — AI skips steps via these evasions. Recognize and reject:
>
> | Evasion                      | Rebuttal                                                      |
> | ---------------------------- | ------------------------------------------------------------- |
> | "Too simple for a plan"      | Simple + wrong assumptions = wasted time. Plan anyway.        |
> | "I'll test after"            | RED before GREEN. Write/verify test first.                    |
> | "Already searched"           | Show grep evidence with `file:line`. No proof = no search.    |
> | "Just do it"                 | Still need TaskCreate. Skip depth, never skip tracking.       |
> | "Just a small fix"           | Small fix in wrong location cascades. Verify file:line first. |
> | "Code is self-explanatory"   | Future readers need evidence trail. Document anyway.          |
> | "Combine steps to save time" | Combined steps dilute focus. Each step has distinct purpose.  |

<!-- /SYNC:rationalization-prevention -->

<!-- SYNC:logic-and-intention-review -->

> **Logic & Intention Review** — Verify WHAT code does matches WHY it was changed.
>
> 1. **Change Intention Check:** Every changed file MUST serve the stated purpose. Flag unrelated changes as scope creep.
> 2. **Happy Path Trace:** Walk through one complete success scenario through changed code
> 3. **Error Path Trace:** Walk through one failure/edge case scenario through changed code
> 4. **Acceptance Mapping:** If plan context available, map every acceptance criterion to a code change
> 5. **Tests Verify Intent:** For test/spec changes, verify tests name the protected business rule or invariant and would fail if that intent breaks.
> 6. **Migration Test Exclusion:** Do not write tests for migration code. Schema/data migrations are one-time execution paths, not core application logic.
>
> **NEVER mark review PASS without completing both traces (happy + error path).**

<!-- /SYNC:logic-and-intention-review -->

<!-- SYNC:bug-detection -->

> **Bug Detection** — MUST ATTENTION check categories 1-4 for EVERY review. Never skip.
>
> 1. **Null Safety:** Can params/returns be null? Are they guarded? Optional chaining gaps? `.find()` returns checked?
> 2. **Boundary Conditions:** Off-by-one (`<` vs `<=`)? Empty collections handled? Zero/negative values? Max limits?
> 3. **Error Handling:** Try-catch scope correct? Silent swallowed exceptions? Error types specific? Cleanup in finally?
> 4. **Resource Management:** Connections/streams closed? Subscriptions unsubscribed on destroy? Timers cleared? Memory bounded?
> 5. **Concurrency (if async):** Missing `await`? Race conditions on shared state? Stale closures? Retry storms?
> 6. **Stack-Specific:** Check the configured language/runtime pitfalls and framework-specific failure modes discovered from local code.
>
> **Classify every finding by consequence, not by fix effort; `SYNC:severity-rubric` is authoritative and this is only a quick reminder:**
> - **CRITICAL → block immediately:** an immediate material security/authorization or safety bypass, secret/PII exposure, destructive action, data loss/corruption, or silent failure on a critical path. A failed binary gate is a separate hard blocker (represented as synthetic CRITICAL by the executable policy), not an ordinary tier judgment.
> - **HIGH → must fix before PASS/merge:** wrong behavior on a supported path, violated business/data invariant, meaningful privacy/authority gap, breaking contract/compatibility change, likely material harm, or missing proof for a behavior-changing fix.
> - **MEDIUM → clear before the current round can pass:** a bounded but consequential edge case, resilience/observability/testability/maintainability gap, credible future defect, or local architectural drift with real impact but no immediate material loss. If the fix needs an owner/product decision, stop and escalate with an explicit follow-up and residual-risk record; that record is not a clean-pass waiver.
> - **LOW → record/defer from round 2 onward:** wording/formatting, minor documentation or convention drift, optional defensive cleanup, or cosmetic refinement with no credible present correctness, security, privacy, authority, availability, or data-integrity impact.
> Assign the highest tier supported by evidence. An observation is not a finding until it names the affected asset/user/data/contract, consequence, evidence, and tier. Effort, annoyance, implementation cost, and proximity to the round cap are never severity inputs. `NOT VERIFIABLE` is an evidence state, not a tier; if the unresolved claim could affect a required behavior or binary gate, keep it blocking until proved or explicitly owner-accepted with residual risk. Failed tests, parity, required artifacts, and security-must-fix checks are binary gates and block independently of severity. For the full decision tree, boundary examples, score mappings, and domain-vocabulary normalization, follow `SYNC:severity-rubric`.

<!-- /SYNC:bug-detection -->

<!-- SYNC:test-spec-verification -->

> **Test Spec Verification** — Map changed code to test specifications.
>
> 1. Identify the project's test/spec format from existing docs, test-case files, BDD feature files, or spec folders.
> 2. Every changed code path MUST ATTENTION map to a corresponding test case/spec (or flag as "needs test case")
> 3. New functions/endpoints/handlers → flag for test spec creation
> 4. Migration files are excluded from TC/test creation; schema/data migrations are one-time execution paths, not core application logic.
> 5. If spec evidence fields exist, verify they point to actual code (`file:line`, not stale references)
> 6. Verify each meaningful test case names the business intent/invariant; flag behavior-only cases that only mirror implementation details.
> 7. Auth/data changes → verify corresponding authorization and data-state test cases exist.
> 8. If no specs exist for a changed path → log the gap and recommend the project's test-spec workflow.
>
> **NEVER skip test mapping.** Untested code paths are the #1 source of production bugs.

<!-- /SYNC:test-spec-verification -->

<!-- SYNC:fix-layer-accountability -->

> **Fix-Layer Accountability** — NEVER fix at the crash site. Trace the full flow, fix at the owning layer.
>
> AI default behavior: see error at Place A → fix Place A. This is WRONG. The crash site is a SYMPTOM, not the cause.
>
> **MANDATORY before ANY fix:**
>
> 1. **Trace full data flow** — Map the complete path from data origin to crash site across ALL layers (storage → backend → API → frontend → UI). Identify where the bad state ENTERS, not where it CRASHES.
> 2. **Identify the invariant owner** — Which layer's contract guarantees this value is valid? That layer is responsible. Fix at the LOWEST layer that owns the invariant — not the highest layer that consumes it.
> 3. **One fix, maximum protection** — Ask: "If I fix here, does it protect ALL downstream consumers with ONE change?" If fix requires touching 3+ files with defensive checks, you are at the wrong layer — go lower.
> 4. **Verify no bypass paths** — Confirm all data flows through the fix point. Check for: direct construction skipping factories, clone/spread without re-validation, raw data not wrapped in domain models, mutations outside the model layer.
>
> **BLOCKED until:** `- [ ]` Full data flow traced (origin → crash) `- [ ]` Invariant owner identified with `file:line` evidence `- [ ]` All access sites audited (grep count) `- [ ]` Fix layer justified (lowest layer that protects most consumers)
>
> **Anti-patterns (REJECT these):**
>
> - "Fix it where it crashes" — Crash site ≠ cause site. Trace upstream.
> - "Add defensive checks at every consumer" — Scattered defense = wrong layer. One authoritative fix > many scattered guards.
> - "Both fix is safer" — Pick ONE authoritative layer. Redundant checks across layers send mixed signals about who owns the invariant.

<!-- /SYNC:fix-layer-accountability -->

<!-- SYNC:source-test-drift-check -->

> **Source/test drift check.** For coding, fix, debug, investigation, test, or review work: when source behavior changes, inspect affected unit/integration/E2E tests and decide from evidence whether tests should change to match intended behavior or the source change is an unintended bug to fix. Do not write tests for migration code; schema/data migrations are one-time execution paths, not core application logic.

<!-- /SYNC:source-test-drift-check -->

<!-- SYNC:ai-mistake-prevention -->

> **AI Mistake Prevention** — Failure modes to avoid on every task:
>
> **Re-read files after context changes.** Context compaction, resume, or long-running work can make memory stale; verify current files before acting.
> **Verify generated content against source evidence.** AI hallucinates APIs, names, claims, and document facts. Check the relevant source before documenting or referencing.
> **Check downstream references before deleting or renaming.** Removing an artifact can stale docs, generated mirrors, configs, and callers; map references first.
> **Trace the full impact chain after edits.** Changing a definition can miss derived outputs and consumers. Follow the affected chain before declaring done.
> **Verify ALL affected outputs, not just the first.** One green check is not all green checks; validate every output surface the change can affect.
> **Assume existing values are intentional — ask WHY before changing OR flagging one as a defect.** Before changing or reporting a constant, limit, flag, cutoff, wording, or pattern, read nearby context and history, the CALLER's ordering, and 2+ sibling call sites of the same convention. A doc stating WHAT without WHY is missing rationale, not proof of a missing guard.
> **Surface ambiguity before acting — don't pick silently.** Multiple valid interpretations require an explicit question or stated assumption with risk.
> **Assert the outcome your system owns, not the intermediate state your infrastructure owns.** When verifying async work, assert the final business state — never the delivery/retry bookkeeping held in shared infrastructure that any co-running process can write. Such a check passes when run alone and flakes the moment anything else shares that infrastructure.
> **Keep shared guidance role-relevant.** Universal guidance must help every receiving skill or agent; code-specific obligations belong only in code-specific protocols.

<!-- /SYNC:ai-mistake-prevention -->

<!-- SYNC:severity-rubric -->

> **Severity Rubric** — Classify every finding by consequence, not by effort, reviewer preference, or how annoying the fix is. One scale applies to every review, skill, agent, workflow, and host so a tier has the same meaning everywhere. Choose the highest credible consequence supported by evidence; do not lower a tier to make a round pass.
>
> **Finding vs observation (required):** An observation becomes a finding only when it names the affected user/system/data/contract, the shipped consequence, the evidence location, and the normalized tier. `INFO`, advice, preference, duplicate wording, or an unsubstantiated concern is not a finding and must not reopen a loop. If the concern might affect a required behavior or gate but evidence is incomplete, emit `NOT VERIFIABLE` with the missing evidence and keep it unresolved; never silently convert uncertainty into LOW.
>
> | Severity | Action | Definition and examples |
> | --- | --- | --- |
> | CRITICAL | Block immediately; escalate | Immediate material risk if shipped: authentication/authorization or safety bypass; secrets/PII exposure; irreversible destructive action; data loss/corruption; or a silent failure on a critical path. A failed binary gate that makes the result untrustworthy is represented as a separate synthetic blocker by the executable policy (not as an ordinary severity judgment). |
> | HIGH | Must fix before PASS/merge | Material correctness or contract risk: wrong behavior on a supported path; violated business/data invariant; meaningful privacy or authority gap; breaking API/schema/compatibility change; likely harm to users/downstream systems; or a missing proof for a behavior-changing fix. |
> | MEDIUM | Must clear the current round; escalate if the fix needs an owner decision | Bounded but consequential risk: an edge case, resilience/observability/testability/maintainability gap, credible future defect, or local architectural drift whose impact is real but not immediate material loss. An explicit follow-up records the escalation/residual risk; it does not make an open MEDIUM a clean pass. |
> | LOW | Record and defer; never open another fix/re-review round from round 2 onward | Non-blocking polish with no credible present correctness, security, privacy, authority, availability, or data-integrity impact: wording/formatting, minor documentation or convention drift, optional defensive cleanup, or a cosmetic/refinement suggestion. |
>
> **Consequence decision tree (apply in order):** (1) Is a binary gate failed? Keep it as a separate hard blocker (the executable helper represents it as synthetic CRITICAL); do not use the ordinary severity label to hide what failed. Otherwise, would shipping permit immediate material security/safety/authority harm, irreversible destruction, data loss/corruption, or a critical-path silent failure? → **CRITICAL**. (2) Otherwise, does a supported path, invariant, public contract, privacy/authority boundary, compatibility promise, or behavior-changing proof fail with material user/downstream impact? → **HIGH**. (3) Otherwise, is there a bounded but consequential edge, resilience, observability, testability, maintainability, or architectural gap with a credible impact? → **MEDIUM**. (4) Otherwise, is the evidence sufficient to show only non-blocking polish with no credible present material impact? → **LOW**. (5) If the evidence needed to choose between steps 1–4 is missing, → **NOT VERIFIABLE**, not LOW. When multiple tiers fit, select the highest credible consequence; effort, implementation cost, reviewer discomfort, frequency alone, and proximity to the round cap never decide the tier.
>
> **Boundary examples (normalize before applying the round predicate):** an auth bypass, exposed secret/PII, destructive command without an authority gate, or failed required test/generation/parity gate is **CRITICAL**; a wrong supported response, broken invariant/API/schema, meaningful privacy/authority defect, or unproven behavior-changing fix is **HIGH**; a bounded retry/timeout/alert/testability gap or credible maintainability drift is **MEDIUM**; a typo, formatting inconsistency, optional cleanup, or cosmetic suggestion proven not to affect present behavior is **LOW**. A missing fact about any of those boundaries is **NOT VERIFIABLE** until evidence or an explicitly documented residual-risk decision exists.
>
> **Classification procedure (required for every finding):** (1) state the affected user, system, data, contract, or gate; (2) assess consequence if the issue ships; (3) assess exposure/likelihood and reversibility/detectability; (4) select the highest tier justified by those facts; (5) cite `file:line` or equivalent evidence and a confidence percentage. Effort, implementation cost, reviewer discomfort, and proximity to the round cap are never severity inputs. `NOT VERIFIABLE` is a pending evidence state, not one of the four tiers and never a LOW escape hatch: if the unresolved claim could affect required behavior, security, privacy, authority, availability, data integrity, or a binary gate, it remains an open evidence blocker until resolved or explicitly owner-accepted with documented residual risk. Classify an item LOW only when evidence supports the absence of credible present material impact.
>
> **Hard-gate rule:** Binary gates (tests, required artifacts, security must-fix checks, generated parity, policy compliance) are not ordinary severity-rated findings. The executable helper records a failed gate as a synthetic CRITICAL blocker solely so one predicate can carry it; the report must still name the gate and failure evidence. A failed gate blocks at every round, including when all ordinary findings are LOW; never disguise a failed gate as LOW.
>
> **Score-based skills** map their numeric scale onto these tiers — do not invent a parallel vocabulary:
>
> - **0-2 criterion scoring** (e.g. production-readiness-review): `0` = CRITICAL/HIGH (criterion unmet, blocks readiness), `1` = MEDIUM (partial, consequential gap), `2` = pass (no finding). If the criterion is only polish, use LOW rather than forcing a `0`.
> - **Two-axis scoring** (e.g. performance-review, impact × likelihood): high impact + high exposure → CRITICAL/HIGH; material impact with bounded exposure → HIGH/MEDIUM; low impact and low exposure → LOW. Record the axes and why the selected tier is the highest credible consequence.
> - **Scorecards / `/20` grades** (e.g. architecture-scalability-review): the aggregate score and verdict band are separate from finding severity. A sub-80 area is evidence to investigate, not an automatic CRITICAL/HIGH/MEDIUM/LOW label; classify each underlying gap by the consequence decision tree and keep advisory score deductions separate from blocking findings.
>
> **Domain-vocabulary normalization (mandatory):** Specialized skills may keep a local reporting vocabulary, but it MUST feed this same four-tier round predicate — never a second severity system:
>
> - `BLOCKED`, `HARD FAIL`, or `FAIL` is a blocking local verdict, not an automatic CRITICAL label. Classify the underlying consequence as CRITICAL when it is an immediate material risk or failed binary gate; otherwise classify it as HIGH or MEDIUM with evidence, while preserving the local block until the owning gate is satisfied.
> - `WARN` is not permission to ignore a finding. Map it to MEDIUM when the gap is consequential, to LOW only when evidence supports no credible present material impact, or upward to HIGH/CRITICAL when the consequence warrants it. `PASS`/compliant is not a finding.
> - UI `P0`/`P1`/`P2`/`P3`/`P4` map to CRITICAL/HIGH/MEDIUM/LOW/LOW respectively as a starting point; override upward only when the evidence shows a higher shipped consequence. A P0/P1 accessibility or task-completion floor remains a blocking gate even when a local UI report calls it a priority rather than a severity.
> - Numeric SRE/readiness or impact/likelihood scores are evidence inputs, not replacement tiers. Emit the score, the consequence, and the normalized CRITICAL/HIGH/MEDIUM/LOW tier together. `INFO`/advisory observations are not findings unless the evidence shows a material consequence.
>
> A finding's tier drives the gate: CRITICAL/HIGH/MEDIUM remain actionable and blocking under the round policy; LOW may be tracked as a follow-up and, from round 2, does not by itself justify another fix/re-review. An owner decision may explain or schedule an open MEDIUM but does not turn it into a clean pass; owner acceptance never makes a failed binary gate pass and must record scope, rationale, and residual risk.

<!-- /SYNC:severity-rubric -->


<!-- SYNC:goal-contract-satisfaction-loop -->

> **Goal Contract Satisfaction Loop** — Persist the user goal in an external file, execute against it, and loop review/fix until every saved required criterion passes or a blocker escalates. Bounded closed loop — NEVER open-ended autonomous exploration.
>
> 1. **Resolve the active goal** (in order): active plan `goal.md` → `plans/goals/{YYMMDD-HHmm}-{slug}/goal.md` → create a new Goal Contract from the current user request (template: `.claude/templates/goal-contract-template.md`).
> 2. **Required sections:** Original Request, Purpose, Success Criteria (checkboxes; mark required vs optional), Constraints, Evidence Required, Iteration Log, Goal Satisfaction matrix.
> 3. **Before work:** read the active goal and map planned work to saved success criteria — execution serves the saved criteria, never chat memory alone.
> 4. **After execution/verification:** append an Iteration Log entry — result, evidence references (`file:line`, command output, report path), remaining gaps.
> 5. **Review gate:** emit a Goal Satisfaction matrix — `| Success Criterion | Evidence | Status |` with PASS/FAIL/BLOCKED. Overall PASS requires every required criterion PASS.
> 6. **Loop rule (retry):** required criterion FAIL → validate the gap is real → fix → re-review only the affected criteria. Stop cleanly when all required criteria PASS.
> 7. **Escalation rule (stop):** two consecutive iterations with no criterion progressing, or a blocker needing user input → mark the criterion BLOCKED with a user-facing reason and escalate. NEVER loop indefinitely.
> 8. **Skip rule:** tiny conversational tasks may skip the goal file ONLY with a recorded one-line reason. User-accepted gate skips are recorded in the goal file with reason and scope.
> 9. **Security:** NEVER store secrets, tokens, credentials, or private customer data in goal files — store evidence references and redact sensitive values.
>
> **Blocked until:** active goal resolved (or skip reason recorded) · saved success criteria read before edits · iteration evidence appended after execution · Goal Satisfaction matrix emitted before any PASS verdict.

<!-- /SYNC:goal-contract-satisfaction-loop -->

<!-- SYNC:trade-off-interrogation-gate -->

> **Trade-Off Interrogation Gate** — ALWAYS ask these THREE questions before ANY verdict, score, finding, or recommendation — about the thing under review AND about every recommendation YOU make. — why: naming a benefit without its price is an endorsement, not a review; the costliest trade-offs are the ones nobody wrote down.
>
> 1. **Is there any trade-off?** Name what it SACRIFICES. "None" / "pure win" is an unfinished analysis, NOT an answer — to claim none, state which dimensions you checked and why each is unaffected: future change cost · complexity · performance/latency · memory/cost · coupling · reversibility · migration burden · operational load · blast radius · security posture · testability · team skill/ramp · delivery time · UX.
> 2. **Is it worth it?** Weigh gain against sacrifice EXPLICITLY — what is gained (with a metric) · what it costs · WHO pays · WHEN it comes due — then emit **WORTH IT / NOT WORTH IT / UNCLEAR**. "Better" with no metric and no cost FAILS this question. NOT WORTH IT → withdraw or replace the recommendation, never keep it as-is.
> 3. **Is the trade-off material enough to CONFIRM WITH THE USER?** A material trade-off is the user's call, never yours. **MATERIAL** when ANY holds: irreversible / one-way door (data migration, public contract, storage format, vendor lock-in) · cost shifted onto someone else (another team, ops/on-call, future maintainer, end user) · one quality attribute traded for another (correctness↔speed, security↔convenience, latency↔cost, simplicity↔flexibility) · a boundary crossed (client↔server tier, service contract, event contract, shared library) · a high-consequence path (auth, money, data integrity, breaking change, High/Medium residual risk) · the worth-it verdict is UNCLEAR.
>
> **MATERIAL → STOP and confirm via `AskUserQuestion` BEFORE the verdict stands** — state the trade-off, both options, what each sacrifices, and your recommendation. **NOT material →** record it inline with a one-line justification and proceed.
>
> **Non-asking execution contexts — ESCALATE BY HANDOFF, never by silence.** `AskUserQuestion` reaches only the main interactive agent: a sub-agent cannot ask the user, and a terminal/verdict-only mode asks nothing by design. When you are running in such a context, the obligation is **redirected, never waived** — do ALL of: (a) complete questions 1 and 2 normally; (b) decide materiality and record it in the Trade-Off Assessment row with `confirmed? = NO — cannot ask from this context`; (c) **name the unconfirmed MATERIAL trade-off explicitly in your returned summary/verdict so the CALLER (or parent orchestrator) escalates it via `AskUserQuestion` on your behalf** — a material trade-off mentioned only inside a report file on disk is NOT a handoff; (d) do not emit an unqualified PASS — mark the verdict as carrying an unconfirmed material trade-off, so the caller's gate stays closed until the user answers. The caller inherits the escalation duty the moment it reads your return.
>
> This carve-out is about **reachability, not convenience**: it applies ONLY where the tool genuinely cannot reach the user (spawned sub-agent, terminal validate/verdict-only mode, non-interactive/headless run). It is NEVER a licence to skip the question, to self-approve a one-way door, or to downgrade materiality because asking is inconvenient — if you CAN ask, you MUST ask.
>
> **Emit a Trade-Off Assessment row** per reviewed decision and per recommendation: `| decision | sacrifices | gain (metric) | who pays, when | WORTH IT/NOT/UNCLEAR | material? | confirmed? |`.
>
> **BLOCKED until:** trade-off named (or dimensions-checked justification given) · worth-it verdict emitted · materiality decided · every MATERIAL trade-off either confirmed with the user OR — in a non-asking context — handed off in the returned verdict for the caller to confirm. A MATERIAL trade-off that is neither confirmed nor handed off can NEVER be PASS, and NEVER gets buried as a Low-severity note.
>
> **NEVER** answer "no trade-off" without checking · decide a material trade-off silently on the user's behalf · let convergence/delivery pressure authorize walking through a one-way door · bundle several material trade-offs into one vague "proceed?".

<!-- /SYNC:trade-off-interrogation-gate -->


<!-- SYNC:evidence-based-reasoning:reminder -->

- **MANDATORY MUST ATTENTION** cite `file:line` evidence for every claim. Confidence >80% to act, <60% = do NOT recommend.
<!-- /SYNC:evidence-based-reasoning:reminder -->

<!-- SYNC:design-patterns-quality:reminder -->

- **MANDATORY MUST ATTENTION** check DRY via OOP (same-suffix → base class), right responsibility (lowest layer), SOLID. Grep for dangling refs after changes.
<!-- /SYNC:design-patterns-quality:reminder -->

<!-- SYNC:complexity-prevention:reminder -->

- **MANDATORY MUST ATTENTION** apply complexity prevention — one business change = one code change. Flag change amplification (>3 edit sites for future change), scattered type-switches, anemic models, primitive obsession, leaked technology through abstractions, shallow modules, un-extracted utility logic (paging/datetime/string/retry → helpers), and logic in the wrong higher layer (downshift to callee/entity/VM). Don't rationalize silent duplication with pure YAGNI.
<!-- /SYNC:complexity-prevention:reminder -->

<!-- SYNC:double-round-trip-review:reminder -->

- **MANDATORY IMPORTANT MUST ATTENTION** execute the review loop (aka **Self-Review Convergence Loop**): review → validate findings → fix validated blocking findings → full re-review. Round 1 ends only with zero findings and the persisted `minRounds` met; from round 2 onward, zero CRITICAL/HIGH/MEDIUM ends the loop once the persisted minimum is met and LOW findings are recorded as deferred. Any newly produced output/judgment gets ≥1 self-review; any new judgment gets ≥1 `/why-review --validate-findings` pass before it is treated as final.
- **MANDATORY** apply the **severity floor**: round 1 exits on zero findings at any severity; **from round 2 the bar is zero CRITICAL/HIGH/MEDIUM — LOW findings are no longer required to be fixed, so a LOW-only round ENDS the loop once the persisted minimum is met.** List every deferred LOW in the report; NEVER re-tier a real CRITICAL/HIGH/MEDIUM down to LOW to reach the exit, and NEVER apply the floor to a binary gate (test-green, security must-fix).
- **MANDATORY** enforce the **round cap of 3 — a ceiling, NEVER a target**: a clean pass ends the loop once the persisted `minRounds` is met (default 1; explicit 2 requires an independent pass), and round 3 completing with CRITICAL/HIGH/MEDIUM still open → **STOP & escalate via `AskUserQuestion`**, never a silent PASS. The 2-repeated-no-progress blocker rule is an earlier exit — escalate at whichever trips first. NEVER loop open-ended.

<!-- /SYNC:double-round-trip-review:reminder -->




<!-- SYNC:rationalization-prevention:reminder -->

- **MANDATORY MUST ATTENTION** follow ALL steps regardless of perceived simplicity. "Too simple to plan" is evasion, not reason.
<!-- /SYNC:rationalization-prevention:reminder -->

<!-- SYNC:graph-assisted-investigation:reminder -->

- **MANDATORY MUST ATTENTION** run at least ONE graph command on key files when graph.db exists. Pattern: grep → graph trace → grep verify.
<!-- /SYNC:graph-assisted-investigation:reminder -->

<!-- SYNC:logic-and-intention-review:reminder -->

- **MANDATORY MUST ATTENTION** verify every changed file serves stated purpose. Trace happy + error paths. Flag scope creep.
<!-- /SYNC:logic-and-intention-review:reminder -->

<!-- SYNC:bug-detection:reminder -->

- **MANDATORY MUST ATTENTION** check null safety, boundary conditions, error handling, resource management for every review.
<!-- /SYNC:bug-detection:reminder -->

<!-- SYNC:test-spec-verification:reminder -->

- **MANDATORY MUST ATTENTION** map every changed function/endpoint to a test. Search for project's test spec format near changed files. Flag coverage gaps, recommend test creation.
<!-- /SYNC:test-spec-verification:reminder -->

<!-- SYNC:translation-sync-check:reminder -->

- **MANDATORY MUST ATTENTION** for multilingual frontend/UI text changes, verify translation updates are present (or explicitly accepted by user as risk) before PASS.
<!-- /SYNC:translation-sync-check:reminder -->

<!-- SYNC:cross-stack-impact-trace:reminder -->

**MUST ATTENTION** FIRST review action — note change context + holistically trace full pipeline of main affected area across client↔server seam (BE→FE forward, FE→BE backward). Verify both tiers still agree on route/DTO/field/type/nullability/auth; any mismatch = BREAKING finding. Skip only for single-tier / docs-only changes (state so).

<!-- /SYNC:cross-stack-impact-trace:reminder -->

<!-- SYNC:cross-service-check:reminder -->

**IMPORTANT MUST ATTENTION** microservices/event-driven: scan producers, consumers, sagas, contracts in task scope. Per touchpoint: owner · message · consumers · risk (NONE/ADDITIVE/BREAKING). Missing consumer = silent regression.

<!-- /SYNC:cross-service-check:reminder -->

<!-- SYNC:fix-layer-accountability:reminder -->

**IMPORTANT MUST ATTENTION** trace full data flow and fix at owning layer, not crash site. Audit all access sites before adding `?.`.

<!-- /SYNC:fix-layer-accountability:reminder -->

<!-- SYNC:critical-thinking-mindset:reminder -->

**MUST ATTENTION** apply critical + sequential thinking — every claim needs appropriate traced evidence (`file:line` for repo/code claims; source URL or artifact section for research, product, content, and docs claims); confidence >80% to act, <60% DO NOT recommend. Anti-hallucination: never present guess as fact, admit uncertainty freely, cross-reference independently, stay skeptical of own confidence.

<!-- /SYNC:critical-thinking-mindset:reminder -->

<!-- SYNC:sequential-thinking-protocol:reminder -->

**MUST ATTENTION** apply sequential-thinking — multi-step Thought N/M, REVISION/BRANCH/HYPOTHESIS markers, confidence % closer; see `/sequential-thinking` skill.

<!-- /SYNC:sequential-thinking-protocol:reminder -->

<!-- SYNC:ai-mistake-prevention:reminder -->

**MUST ATTENTION** apply AI mistake prevention — verify generated content against evidence, trace downstream references before deleting or renaming, verify all affected outputs, re-read files after context loss, and surface ambiguity before acting.

<!-- /SYNC:ai-mistake-prevention:reminder -->

<!-- SYNC:category-review-thinking:reminder -->

- **MANDATORY** Derive review categories from file language + directory semantics + change nature; create a sub-task per category.
- **MANDATORY** Derive each category's concerns from first principles with `file:line` evidence — never a fixed checklist.

<!-- /SYNC:category-review-thinking:reminder -->

<!-- SYNC:task-tracking-external-report:reminder -->

- **MANDATORY** Bootstrap task tracking before target work; transition one task at a time.
- **MANDATORY** Persist plan/review findings to `plans/reports/` incrementally and synthesize from disk.

<!-- /SYNC:task-tracking-external-report:reminder -->

<!-- SYNC:project-reference-docs-guide:reminder -->

- **MANDATORY** Before investigating, planning, or coding, read `docs/project-config.json` (the project map: modules/paths, run-commands, conventions, architecture/workflow rules) + the required project-reference docs, and cite `Reference docs read: ...`.
- **MANDATORY** Load detail just in time immediately before the first target read/grep/edit/test; hooks may provide a pointer, but a hook event or prior turn is never evidence that the current files were read.
- **MANDATORY** Always include `lessons.md`; project config + conventions override generic framework defaults.
- **MANDATORY** If project config, root instruction files, or any required reference doc is missing or stale, auto-run `/project-init` or the narrow lower-level route before ordinary project-specific work. On compaction, resume, delegation, or a context change, re-read the required docs and restate the route before continuing.

<!-- /SYNC:project-reference-docs-guide:reminder -->

<!-- SYNC:end-to-start-debugger-trace:reminder -->

**IMPORTANT MUST ATTENTION** debugger trace gate: for non-trivial bug/fix/investigation/review work, start at the observed final output and trace backward through reader -> storage/projection -> writer -> consumer/job -> producer/trigger. Enumerate all feeder paths and hypotheses before fixing. **BLOCKED until** trace, hypothesis matrix, owning fix layer, and forward convergence proof exist.

<!-- /SYNC:end-to-start-debugger-trace:reminder -->

<!-- SYNC:nested-task-creation:reminder -->

- **MANDATORY** Parent workflow rows do not replace child phase tracking; expand phases and link the parent when nested.
- **MANDATORY** Orchestrators pre-expand child skill phases before invocation; use `[N.M] /skill-name — phase` prefixes and one-`in_progress` discipline.

<!-- /SYNC:nested-task-creation:reminder -->

<!-- SYNC:goal-contract-satisfaction-loop:reminder -->

- **MANDATORY** Resolve the active Goal Contract BEFORE work (active plan `goal.md` → `plans/goals/{YYMMDD-HHmm}-{slug}/goal.md` → create from current request) and read saved success criteria before editing.
- **MANDATORY** Append iteration evidence after execution; emit a Goal Satisfaction matrix (PASS/FAIL/BLOCKED) before reporting PASS; loop on validated FAIL; escalate repeated no-progress or blockers. NEVER store secrets in goal files.

<!-- /SYNC:goal-contract-satisfaction-loop:reminder -->

<!-- SYNC:systematic-review-batching:reminder -->

- **MANDATORY** Large changeset → batch by size cap (≤8 files OR ≤2000 diff-lines), one parallel sub-agent per batch; never review many files one-by-one.
- **MANDATORY** > 6 categories OR > 40 files → add the hierarchical synthesis tier; each concern-synthesizer emits cross-concern interaction candidates and the orchestrator runs the cross-concern pass before concluding.

<!-- /SYNC:systematic-review-batching:reminder -->

<!-- SYNC:severity-rubric:reminder -->

- **MANDATORY** Classify every finding Critical/High/Medium/Low by consequence using the affected asset, shipped impact, exposure, reversibility, evidence location, and confidence; Critical/High/MEDIUM remain actionable under the round bar, while LOW is recorded/deferred from round 2 onward.
- **MANDATORY** Keep binary gates separate from severity: a failed test, security must-fix, required artifact, or parity check blocks at every round and is never relabeled LOW.
- **MANDATORY** Score-based skills (sre 0-2, perf two-axis) map onto the same four tiers — no parallel severity vocabulary.

<!-- /SYNC:severity-rubric:reminder -->


<!-- PROMPT-ENHANCE:STEP-TASK-CLOSING:START -->

## Prompt-Enhance Closing Anchors

**IMPORTANT MUST ATTENTION** follow declared step order for this skill; NEVER skip, reorder, or merge steps without explicit user approval
**IMPORTANT MUST ATTENTION** for every step/sub-skill call: set `in_progress` before execution, set `completed` after execution
**IMPORTANT MUST ATTENTION** every skipped step MUST include explicit reason; every completed step MUST include concise evidence
**IMPORTANT MUST ATTENTION** if Task tools unavailable, maintain an equivalent step-by-step plan tracker with synchronized statuses

<!-- PROMPT-ENHANCE:STEP-TASK-CLOSING:END -->

<!-- SYNC:trade-off-interrogation-gate:reminder -->

- **MANDATORY MUST ATTENTION ALWAYS ASK THE 3 TRADE-OFF QUESTIONS** — on the thing under review AND on every recommendation you make: (1) **is there any trade-off?** name what it SACRIFICES (change cost · complexity · perf · coupling · reversibility · migration · ops load · blast radius · security · testability · delivery time · UX) — "none"/"pure win" is an unfinished analysis, so state the dimensions checked; (2) **is it worth it?** gain (with a metric) vs cost, WHO pays, WHEN → emit **WORTH IT / NOT WORTH IT / UNCLEAR**; NOT WORTH IT → withdraw or replace it; (3) **is it material enough to confirm with the user?** irreversible/one-way door · cost shifted onto another team/ops/maintainer/user · one quality attribute traded for another · a tier/service/event/library boundary crossed · auth/money/data-integrity/breaking-change/High-or-Medium-risk path · verdict UNCLEAR → **STOP and confirm via `AskUserQuestion` BEFORE the verdict**.
- **MANDATORY** A MATERIAL trade-off with no user confirmation can NEVER be PASS; NEVER bury one as a Low-severity note, NEVER decide it silently, and NEVER let delivery or convergence pressure authorize a one-way door. — why: an un-walked-back one-way door is the user's call to make, not the reviewer's.
- **MANDATORY — non-asking contexts escalate BY HANDOFF, never by silence.** `AskUserQuestion` reaches only the main interactive agent: a sub-agent cannot ask the user, and a terminal/verdict-only mode asks nothing by design. There the duty is REDIRECTED, not waived — still name the trade-off, still decide materiality, record `confirmed? = NO — cannot ask from this context`, **state the unconfirmed MATERIAL trade-off in your RETURNED verdict/summary so the CALLER escalates it** (a note only in an on-disk report is not a handoff), and never emit an unqualified PASS. Applies ONLY where the user is genuinely unreachable (spawned sub-agent, terminal validate mode, headless run) — if you CAN ask, you MUST ask.

<!-- /SYNC:trade-off-interrogation-gate:reminder -->

<!-- SYNC:parallel-subagent-dispatch -->

> **Parallel Sub-Agent Dispatch** — Plan parallelism the moment a task breakdown exists, BEFORE executing it — running provably independent tasks sequentially wastes wall-clock. Applies to every multi-step job: workflow steps, planning, batch updates, investigation, research, scans, reviews, doc sync. **Plan execution is metadata-gated, NEVER default-parallel** — fan-out follows ONLY what the plan declares (`PAR`/`SEQ` tags + per-phase write set); an untagged plan runs sequentially — why: a derived write set cannot see cascade or generated writes.
>
> 1. **Tag every task `PAR` or `SEQ`.** `PAR` = inputs exclude every pending task's output AND write set disjoint from every other `PAR`. Else `SEQ` — MUST ATTENTION name the dependency forcing it.
> 2. **Group `PAR` into waves.** No edge between members. Two writers of one file NEVER share a wave. Read-only work (search, investigation, review, research) parallelizes freely.
> 3. **Declare before dispatch:** `Parallel plan: wave 1 = [...] · wave 2 = [...] · SEQ = [...] (reason)`.
> 4. **Spawn each wave in ONE message** — every `Agent` call in one response, NEVER dripped per turn. Route each task to its specialist (`.claude/skills/shared/sub-agent-selection-guide.md`); NEVER `code-reviewer` as catch-all.
> 5. **Brief each sub-agent self-contained:** goal · scope + owned files · reference docs · return contract (summary + `Full report:` path, per SYNC:subagent-return-contract) · incremental persistence to `plans/reports/` (per SYNC:incremental-persistence).
> 6. **Barrier per wave.** Advance ONLY after EVERY member returns (a skipped conditional counts as returned). Merge, mark each task completed/skipped, THEN dispatch the next wave. Mutating steps wait for the barrier.
> 7. **One level deep.** A dispatched sub-agent executes its own brief; further fan-out stays the orchestrator's job unless that agent's `.claude/agents/*.md` definition authorizes it.
>
> **NEVER parallelize:** tasks sharing a write target · a task consuming a pending task's output · trivial single-file work (dispatch overhead > gain) · an order a skill or workflow explicitly fixes · gates awaiting user approval.
>
> **Blocked until:** MUST ATTENTION every task tagged PAR/SEQ with a named reason per SEQ · waves declared + write-set disjointness checked · each wave spawned in ONE message · barrier honored before the next wave.

<!-- /SYNC:parallel-subagent-dispatch -->

<!-- SYNC:parallel-subagent-dispatch:reminder -->

- **MANDATORY** After planning tasks, tag each PAR/SEQ and spawn every PAR wave as parallel sub-agents in ONE message — default parallel for workflows, batch updates, investigation, research, reviews; plan execution fans out ONLY on what the plan declares.
- **MANDATORY** Disjoint write sets per wave · all-return barrier before the next wave · specialist routing · sub-agents NEVER fan out further unless their own agent definition authorizes it.

<!-- /SYNC:parallel-subagent-dispatch:reminder -->

<!-- SYNC:project-protocol-overlay -->

> **Project Protocol Overlay** — Before executing this skill, resolve any PROJECT overlay rules layered onto it: match this skill's name against the `Target` column of the project's skill-protocol index (`docs/project-reference/skill-protocols-reference.md` by default; a `referenceDocs` entry in `docs/project-config.json` overrides the path), taking the most specific matching tier ONLY — exact name > glob > `*`. **That precedence orders overlays against EACH OTHER, never against this skill.** Read ONLY the matched bodies, resolved as `<protocols-dir>/<Name>.md`; a row's Body link is display text, never a read path. A matched body that is missing or malformed is REPORTED and skipped — never reconstructed from the index Description. No index, or no match -> proceed with no overlay, silently. Full contract: `.claude/skills/project-skill-protocol/references/registry.md`.
>
> Overlays are **ADDITIVE ONLY**: they ADD rules on top of this skill's own protocol and NEVER replace, override, disable, or reinterpret a rule it already states — removing every overlay must return this skill to exactly its documented behavior. An overlay is a BRIEF, not an authority escalation: it can NEVER waive a workflow gate, git discipline, a review gate, or a user-confirmation gate. A genuine overlay-vs-skill conflict, or two equally-specific overlays that directly contradict -> surface both to the user; NEVER resolve silently.

<!-- /SYNC:project-protocol-overlay -->

<!-- SYNC:project-protocol-overlay:reminder -->

**MUST ATTENTION** resolve project protocol overlays for this skill BEFORE executing — most specific matching tier only (exact > glob > `*`, which ranks overlays against each other, NEVER against this skill), read only matched bodies at `<protocols-dir>/<Name>.md`; a missing or malformed body is reported, never reconstructed. Overlays are ADDITIVE ONLY (they never replace this skill's own rules) and are a brief, NEVER an authority escalation; an equal-specificity contradiction goes to the user.

<!-- /SYNC:project-protocol-overlay:reminder -->

## Closing Reminders

**IMPORTANT MUST ATTENTION Goal:** Ensure reviewed code is correct, easy to change, convention-aligned, and verification-backed before acceptance or handoff — via receiving feedback with verification (not performative agreement), requesting targeted systematic reviews through the code-reviewer subagent, and enforcing verification gates before completion claims.

**IMPORTANT MUST ATTENTION — Main steps:** create report → graph/blast-radius + full-pipeline trace → detect risks/plan/surfaces → file-by-file review → holistic review → final result → validate findings before fixes → full re-review after fixes; current/branch diff requests route to `changes-review`.

**MUST ATTENTION Protocols in force (concise digest of the SYNC/shared blocks this skill carries — each line is a signpost to its canonical body above; the body governs):** NEVER act on the digest alone; ALWAYS read the cited canonical block when its trigger fires.

- **Systematic Batching:** Large changeset → size-capped parallel batches, reduce.
- **End To Start Debugger Trace:** Trace observed end-state backward before fixing.
- **Graph Assisted Investigation:** Run graph command on key files first.
- **Category Review Thinking:** Derive category concerns from first principles.
- **Subagent Return Contract:** Sub-agents return summary only, report on disk.
- **Nested Task Creation:** Expand child phases under workflow row.
- **Project Reference Docs Guide:** Read project docs before target work.
- **Task Tracking External Report:** Bootstrap tasks; persist findings incrementally.
- **Critical Thinking Mindset:** Traced proof per claim; never guess.
- **Sequential Thinking Protocol:** Multi-step Thought N/M with confidence closer.
- **Evidence Based Reasoning:** Cite `file:line`; confidence >80% to act.
- **Design Patterns Quality:** DRY via OOP, lowest layer, SOLID.
- **Complexity Prevention:** One business change = one code change.
- **Double Round Trip Review:** Review → validate → fix current-round blocking findings → re-review until the current severity bar is clear (round 1: zero findings; round 2+: zero CRITICAL/HIGH/MEDIUM, LOW deferred).
- **Fresh Context Review:** Re-review via fresh-memory sub-agents post-fix.
- **Review Protocol Injection:** Embed all 11 protocol bodies verbatim.
- **Rationalization Prevention:** Reject step-skipping evasions; show evidence.
- **Logic And Intention Review:** Changed code matches stated purpose.
- **Bug Detection:** Check null/boundary/error/resource every review.
- **Test Spec Verification:** Map every changed path to a test.
- **Fix Layer Accountability:** Fix at owning layer, not crash site.
- **Source Test Drift Check:** Source change → reconcile affected tests.
- **AI Mistake Prevention:** verify generated content against evidence, trace downstream references, verify all affected outputs, re-read after context loss, surface ambiguity.
- **Severity Rubric:** Classify by consequence; Critical/High block PASS.

- **MANDATORY** Nested Task Expansion Contract — when invoked inside a workflow, STILL expand internal phases via `TaskCreate` with `[N.M] /skill-name — phase` prefix and `TaskUpdate(parentTaskId, addBlockedBy: [childIds])` linkage. Workflow row is container, not substitute.
- **MANDATORY** break work into small todo tasks using `TaskCreate` BEFORE starting
- **MANDATORY** validate decisions with user via `AskUserQuestion` — never auto-decide
- **MANDATORY** add final review task to verify work quality
- **MANDATORY MUST ATTENTION** search for project-specific reference docs BEFORE reviewing (coding standards, architecture, test conventions)
- **MANDATORY MUST ATTENTION** Phase 0: detect change type FIRST — route auth/perf files to specialized sub-agents before general review
- **MANDATORY MUST ATTENTION** run `/why-review` after completing this review to validate design rationale, alternatives considered, and risk assessment
- **Parallel Sub-Agent Dispatch:** Tag tasks PAR/SEQ, group PAR into disjoint-write-set waves, spawn each wave in ONE message, barrier before advancing.

**[TASK-PLANNING]** Before acting, analyze task scope and systematically break it into small todo tasks and sub-tasks using TaskCreate.

---

> **Closing reminder — Easy to Change is the success metric.** Every finding,
> test, refactor, and abstraction must answer one question: _does this make
> the next change cheaper or more expensive?_ If it doesn't reduce future
> change cost, reject it. Coupling, hidden state, duplicated knowledge, and
> unclear intent are the real enemies — call them out by name.
**Anti-Rationalization:**

| Evasion | Rebuttal |
| ------- | -------- |
| "Purpose obvious" | Anchor it anyway — primacy/recency keeps outcome active through long prompts. |
| "Existing reminders enough" | Echo Goal in Closing Reminders — bottom anchor prevents drift. |
| "Skip evidence for prompt edits" | Cite changed file evidence and verify no stale protocol text remains. |