Claude Skills by machenjie

`analysis-agent`/`task-agent`/`review-agent`: use when API/DTO, domain, persistence, event, generated, null/default, or mapper boundaries may leak; skip when models stay isolated.

`analysis-agent`/`review-agent`: use when module ownership, public surface, dependency direction, cycles, or shared state changes; skip owner-internal placement.

`analysis-agent`/`task-agent`/`review-agent`: use when HTTP/TLS/DNS, proxies, ingress, headers, timeout chains, WebSocket, or gateways change; skip without network-edge impact.

Use when Node.js event-loop, stream, Buffer, process, Worker, async-context, cancellation, or handle semantics change. Skip TypeScript, browser, build, and business rules.

`analysis-agent`: use when scope, exclusions, deferred decisions, version limits, or assumptions need explicit boundaries; skip when no task-local non-goal decision is needed.

`task-agent`: use when document, key-value, wide-column, or graph storage changes access, partitioning, consistency, or evolution; skip vendor-only mentions and unchanged storage.

`analysis-agent`/`task-agent`/`review-agent`: primary-Skill-selected for logs, metrics, traces, alerts, SLI/SLO, or diagnostics; never task owner; skip without signal impact.

`analysis-agent`/`task-agent`/`review-agent`: use offline to find validation, review, repair, or closure gaps; skip without sequence evidence; never infer live performance.

`analysis-agent`/`task-agent`/`review-agent`: use for offline authority, pending operations, conflicts, cursors, transfers, or reconnect recovery; skip online-only work.

`analysis-agent`/`task-agent`/`review-agent`: use when package managers, lockfiles, additions, upgrades, removals, licenses, or supply-chain risks change; skip unrelated work.

`analysis-agent`/`task-agent`: use when page components, props, state ownership, reuse, or decomposition boundaries change; skip when no component-structure decision exists.

`analysis-agent`/`task-agent`/`review-agent`: use when latency, throughput, bundle, memory, CPU, query, rendering, or resource cost needs a budget; skip without performance risk.

`analysis-agent`/`task-agent`/`review-agent`: use for tenant isolation, privilege-escalation, subject-resource-action policy, or enforcement changes; skip authentication-only work.

`analysis-agent`/`task-agent`/`review-agent`: use when diff, behavior, validation, review, or re-review may drift from an accepted plan; skip without a closure consistency need.

`analysis-agent`/`task-agent`/`review-agent`: use for PowerShell pipeline, error, native, encoding, remoting, credential, provider, module, or idempotency decisions.

`analysis-agent`/`task-agent`/`review-agent`: use when personal-data purpose, retention, deletion, sharing, telemetry, or provider handling changes; skip legal-only work.

`task-agent`/`review-agent`: use when CPU, memory, I/O, database, network, rendering, or cost needs measured bottleneck evidence; skip without a profiling need.

`analysis-agent`/`task-agent`/`review-agent`: use for structure, workspace or package bootstrap, config, scripts, or quality gates; skip without an initialization decision.

`analysis-agent`/`task-agent`: use when a prototype resolves hierarchy, interaction, failure, responsive, or accessibility uncertainty; skip visual-only polish.

`analysis-agent`/`task-agent`/`review-agent`: use when Python changes cross import, async, resource, typing, mutability, or serialization boundaries; skip tool-only work.

`analysis-agent`/`review-agent`: use when splitting, merging, moving, or deleting code while preserving behavior; skip behavior-changing implementation.

`analysis-agent`/`task-agent`/`review-agent`: use for recurrence guards on known defects, incidents, or escaped failures; skip speculative risk without a prior failure mechanism.

`task-agent`: use when physical relational schema or database-enforced integrity changes; skip conceptual-model, repository-only, or unchanged relational-storage work.

`analysis-agent`/`task-agent`/`review-agent`: use for release identity, compatibility, exposure, stop, rollback, or recovery; skip pipeline-only and no-release work.

`analysis-agent`/`task-agent`/`review-agent`: use when repeated failure needs a new hypothesis or proof path; skip an initial failure with verified cause and a different action.

`analysis-agent`/`task-agent`/`review-agent`: discover source of truth, candidate owner, and change surface before planning; hand known candidates to impact inspection.

`analysis-agent`/`task-agent`/`review-agent`: use after a candidate owner/change surface is known and consumers, tests, contracts, or generated impact need bounded proof.

`task-agent`: use for repository methods, query behavior, record mapping, visibility, errors, or transaction participation; skip schema, migration, DTO, and domain-rule work.

`analysis-agent`: use when goals, constraints, assumptions, or missing facts need blocking, non-blocking, and safe-assumption decisions; skip when requirements are clear.

`analysis-agent`: use when raw requests need behavior, actors, scope, non-goals, constraints, deliverables, acceptance, or test traceability; skip when structure already exists.

`analysis-agent`/`task-agent`: use when routes, guards, redirects, deep links, breadcrumbs, back behavior, stale links, or recovery change; skip when navigation is unaffected.

`analysis-agent`/`task-agent`/`review-agent`: use when Rust changes cross ownership, unsafe/FFI, panic, cancellation, or Send/Sync boundaries; skip tool-only work.

`analysis-agent`: use when a request needs normal, failure, edge, abuse, recovery, or operational scenarios; skip when no scenario-decomposition decision exists.

`analysis-agent`/`task-agent`/`review-agent`: use when SDK/library APIs, semver, compatibility, deprecation, or migration changes; skip without library-contract impact.

`analysis-agent`/`task-agent`/`review-agent`: use for search relevance, freshness, analytical grain, metrics, or rebuilds; skip bounded source queries and operational telemetry.

`analysis-agent`/`task-agent`/`review-agent`: use when secrets, KMS/env config, rotation, logging, redaction, or rollback changes; skip non-sensitive config.

`analysis-agent`/`task-agent`/`review-agent`: use for cross-cutting source-backed judgment only when no narrower capability owns it; skip trivial or narrower-owned work.

`analysis-agent`/`task-agent`: use when a use case coordinates authorization, domain work, transactions, or external effects; skip transport, storage, and rule-only work.

Use with task-agent, review-agent, or analysis-agent for task-local Shell/CLI safety and rerun behavior. Do not use without that decision or as task owner.

`analysis-agent`/`task-agent`/`review-agent`: use when authoring SKILL.md, triggers, decisions, references, registry, routing, or validation; skip when no Skill content changes.

`analysis-agent`/`task-agent`/`review-agent`: use when Skill changes need baseline/treatment evidence for routing, defects, or regressions; skip without an efficacy claim.

`analysis-agent`/`task-agent`/`review-agent`: use when feasible solutions differ in hard constraints, total-change cost, or reversibility; skip when no material option remains.

`analysis-agent`/`task-agent`/`review-agent`: use when SQL changes affect NULL, cardinality, constraints, locks, precision, or dialect semantics; skip tool-only work.

`analysis-agent`/`task-agent`/`review-agent`: use when lifecycle states, transitions, guards, or terminal states need modeling; skip when no state-machine decision exists.

`task-agent`: use when local, server, global, derived, optimistic, persisted, form, auth, or cache state ownership changes; skip when state boundaries are unaffected.

`analysis-agent`/`task-agent`/`review-agent`: use for Swift ownership, concurrency, type, Objective-C interop, SwiftUI state, or package-boundary decisions.

Select exact repository-defined validation commands and coverage after a proof strategy exists; use when entrypoint choice or coverage is unresolved.

Select facts, artifacts, Layer 3 Skills, and References when decision context is stale, redundant, uncertain, or over budget; skip transfer and sufficient current context.

Analyze candidate task nodes, dependencies, collisions, and critical path before a consumer accepts or rejects a final DAG.

Select evidence-bearing context for a downstream consumer after work when claims, artifacts, validation, unresolved decisions, or omissions must cross a boundary.