Claude Skills by OutlineDriven

Use when changing the wording of a rule in a skill, prompt template, or agent instruction where the change is meant to alter model output. Produces a scored results table comparing each variant against a no-guidance control, with every match hand-verified.

Use when skill or knowledge activation is unreliable, or setup reaches activation reliability. Configures a project-scoped, marker-gated hook and measures affected evals before and after. Don''t use for remote, credential, publish, deploy, or irreversible changes.

Use when a human invokes the release gate to prove all 28 ODIN plugins share one canonical version. Emits a per-file comparison and exits non-zero on mismatch. Don''t use to edit release metadata or for remote, credential, publish, deploy, or irreversible changes.

Use when asked to create an MCP server to integrate an API or service in Python or TypeScript. Produces a server with typed tools, tests, and a read-only evaluation suite. Don''t use for remote, credential, publish, deploy, or irreversible changes.

Use when plan-mode enforcement hooks intercept a plan review or the user runs /plan-review-tune to tune which plan-review questions fire. Persists the tuned question flow to a local question-registry and hook configuration. Not for auditing a plan against code — use plan-review.

Use when a user wants to define patterns the agent should not do; result is structured project docs or agent-rule files listing each prohibited pattern. Don''t use for remote, credential, publish, deploy, or irreversible changes.

Use when asked to improve, optimize, rewrite, tune, or port prompts, or build prompt evals. Returns a shorter, reliable prompt validated on holdout cases with target, success criteria, external context, and adapter notes.

Use when a completed task needs reflection on the invoked skills to propose and apply approved improvements, with explicit human approval before any edit. Not for remote, credential, publish, deploy, or irreversible changes.

Use when asked to review a plugin for marketplace readiness through a read-only audit of every published quality gate. Returns a sectioned pass/fail report with a submission recommendation. Not for reviewing a PR or code diff — use review.

Use when installing a third-party automation pack or template repository safely and deterministically. Fetches the source archive, verifies every file against an expected manifest of hashes, applies required transformations using strict unique anchors, checks for destination conflicts, and atomically moves files to the target with a saved rollback manifest. Not for unattended runs; requires explicit human invocation.

Use when asked to configure pstack role-to-model choices so every role maps to an available model. Writes one user-level pstack model rule file. Not for remote, credential, publish, deploy, or irreversible changes.

Use when the user asks to set up or configure the agent environment from a fresh or existing clone. Configures credentials, placeholders, and the virtual environment with explicit human confirmation at each step. Not for automated or unattended runs.

Use when a skill change is about to ship and must pass a release gate. Validates gate inputs fail-closed, runs activation and regression cases against an explicit baseline, measures context cost against an explicit ceiling, and emits PASS or FAIL with one issue per failure. Not for scoring without gating — use skill-benchmark.

Use when the user runs /skill-benchmark to score agent skills via LLM judges with baseline comparison, regression detection, and trend analysis, or to compare candidate models on a shared task set in a ranked table with per-model spend tracking. Not for release gating — use skill-benchmark-gate.

Use when a user wants agent setup graded from conversation history. Produces an HTML report with 0-10 scores, evidence-cited findings, and ranked suggestions. Not for skill fixing — use skill-improver; not for security scanning — use skill-scanner.

Use when the model is creating or refactoring a skill, faces a SKILL.md over 300-400 lines, or confuses triggering with disclosure. Produces a justified split-or-monolith decision with pointer hygiene and an architecture-eval plan. Not for skill authoring, use writing-skills.

Use when a user asks to scan, audit, or validate a skill for security issues. Runs static analysis on skill files and returns a risk-level assessment with an install recommendation. Not for source-code or infrastructure review—use security-review—or for remote-system changes.

Use when a publisher wants a discoverable, integrity-protected agent-skill catalog served from a domain they control. Builds the discovery index with SHA-256 digests, deterministic flat archives, and verified install commands, then hands the publisher a post-deploy verification checklist. Not for remote mutation or deployment — the publisher deploys the output tree.

Use when the user asks to discover, generate, refresh, or validate skills from the Trail of Bits Testing Handbook or appsec.guide. Not for tasks that require source or remote-system changes.

Use when the user runs /toolchain-health and wants a trustworthy green/yellow/red verdict on the installed toolchain with ranked repairs. Runs the project diagnostic tools read-only under a stated timeout, scores five categories against a threshold table, and presents a weighted composite verdict with a per-category dashboard. Not for tasks that require source or remote-system changes.

Use when asked to audit validation that looks too clean or self-confirming. Identify each independence failure that applies, prescribe its fix, and audit the auditor''s own evidence. Not for checking eval leakage — use evaluation-leakage-audit.

Use when a proven watch pattern should become a reusable harness artifact that replays the proven watch behavior with configurable inputs. Don''t use for remote, credential, publish, deploy, or irreversible changes.

Use when the user runs /workspace-unfreeze on a frozen path to remove the freeze lock marker so the path is editable again. Resolves the marker from an explicit path or the freeze convention, previews the consequence, and deletes only that marker. Not for automated or unattended runs; requires explicit human invocation.

Use when asked to author or restructure any agent-consumed document so the agent routes and executes predictably. Produces a self-contained document with no stale duplication or unreachable pointers. Not for skill-specific mechanics like invocation choice — use writing-skills.

Use when a SKILL.md or AGENTS.md/CLAUDE.md is being authored, refactored, ported, or upgraded, or the user asks to write a skill, improve one, or fix unreliable skill firing. Not for general agent-consumed documents, use writing-for-agents.

Use when the user wants a clean Xcode rebuild by deleting DerivedData and build artifacts. Do not use for diagnosing a specific build error; that is ios-build-fix.

Use when asked to run /ios-build-fix to fix a failing iOS build or UI behavior through the debug bridge. Not for a clean rebuild — use ios-build-cleanup.

Use when the user runs /ios-device-qa to drive a real iPhone over USB through a debug-bridge daemon and return a device QA report with verified interactions. Do not use for remote, credential, publish, deploy, or irreversible changes.

Use when /xcode-project-sync must regenerate an Xcode project from project.yml through gstack templates. Verifies preconditions, checks freshness, regenerates, reviews the diff, and confirms the build. Not for handwritten Swift files, edits outside the generated xcodeproj, a missing project.yml, or simulator testing; use xcode-simulator-testing for the latter.

Use when asked to run /xcode-simulator-testing with a scheme name or current to build and launch an iOS app in a simulator and capture per-surface evidence. Not for project regeneration — use xcode-project-sync.

Use when the user requests a deep dive, exploratory analysis, or data analysis on BigQuery. Produces a dated analyses dir with an approved plan, cohorts, linked SQL and visualization artifacts, and TLDR. Read-only BigQuery warehouse queries are within authority; not for credential, publish, deploy, or irreversible changes.

Use when a human-curated dbt model index must guide BigQuery SQL for a warehouse question. Emits a query with the correct fully-qualified model, grain, standard filters, partition bounds, and cost controls. Not for discovering undocumented models or executing warehouse changes.

Use when non-trivial code needs designing before implementation, or an existing codebase needs its architecture improved: surveying for shallow modules, locating a seam, deepening a structure, hardening with guardrails, or making code testable. Also handles surveying for shallow modules, locating seams, deepening structures, hardening with guardrails, and making code testable. Not for architecture diagrams, deploy, remote, credential, or irreversible changes.

Use when asked to validate a web app, CLI, API, or generated artifact against a source-blind behavior contract. Produces a structured pass/fail/blocked/out-of-scope report with anti-cheat probes and redacted evidence. Not for source or remote-system changes.

Use when asked to determine what a change could break before it ships. Returns confirmed risks, cleared items, and the cheapest pre-merge test that catches the real bug. Not for remote, credential, publish, deploy, or irreversible changes.

Use when bloated code needs clean re-derivation, or the user says "this module is bloated" or "break it and rebuild". Classifies old behavior as essential or residue, cuts residue, and leaves the verifier green. Not for untracked data or changes without VCS rollback.

Use when a candidate change must be produced without touching the working tree; an ephemeral worktree runs the declared command and returns a binary-safe patch plus its exit code. Not for remote, credential, publish, deploy, or irreversible changes.

Use when a requested sweep monitors CI failures over a bounded attempt window. Returns each root cause reproduced or classified non-actionable with any minimal verified patch as a proposal. Not for classifying one failure without patching — use classify-ci-failure.

Use when asked to run /clean-clean-cut to cut accumulated records and residue under an explicit destructive gate with version-control recovery. Don''t use for untracked targets, changes without a version-control rollback, or git branch/worktree cleanup — use git-cleanup.

Use when a user wants to identify the true sources of complexity qualitatively before counting metrics. Returns a ranked complexity-source report with root causes before any metric is counted. Read-only; no source or remote mutation.

Use when asked to reproduce, profile, or verify CLI/TUI behavior. Produces a deterministic transcript or profile proof with session cleanup. Not for CLI design advice — use cli-for-agents.

Use when asked to verify or reproduce browser or Electron UI behavior by capturing before and after evidence with no leftover processes. Not for remote, credential, publish, deploy, or irreversible changes.

Use when asked to remove old code, migrate consumers, or decide whether to maintain or sunset a system. Produces a proven replacement, migrated consumers, removed obsolete tracked code, and rollback evidence. Not for untracked data or changes without VCS rollback.

Use when asked to design a module interface, seam, or testable boundary. Produce compared alternative interface designs using explicit deep-module vocabulary and deliberately placed seams. Not for UI direction picking — use design. No source or remote-system changes.

Use when code mixes units, fixed-point precisions, scaling factors, rates, prices, shares, or conversions and needs dimensional consistency validated. Annotates every in-scope file with unit comments, derives a units vocabulary, and reports confirmed and refuted mismatches. Not for type-level unit modeling — use type-driven.

Use when reconciling an API endpoint''s generated OpenAPI schema and declared response types with its actual runtime response. Adds or fixes schema decorators, reuses canonical types, migrates legacy path definitions, and validates the spec locally. Not for general API documentation: use docs-and-adrs.

Use when asked to prove coverage, find missing cases, or enumerate a state, decision, requirement, or behavior space. Enumerates every cell as covered, gap, or deferred with an executed check per cell, and emits a coverage manifest with a tally. Not for round-based questioning — use askme; not for testing one property — use property-test-authoring.

Use when asked to run a performance campaign against a measured floor: rebuild hot paths from their floor, grade cold paths, and land each target as an atomic commit with a proven win. Not for hypothesis-only analysis without mutation — use fastopt.

Use when a user wants to optimize estimated hot paths and complexity-neutral cold paths while refusing complexity theater. Emits hypotheses and benchmark designs. Not for measured-floor rebuilds — use extremely-optimize; not for hypothesis-only — use fastopt.

Use when a user wants to optimize suspected hot paths without waiting for benchmarks: label every cost observation as a hypothesis and emit benchmark designs, with no mutation or speed claim. Not for measured-floor rebuild campaigns — use extremely-optimize.