
Verify a PR against a task spec — extract acceptance criteria, decide whether rollout expectations matter, audit feature-flag behavior only when the spec or diff requires it, run full code review, and audit completeness. Use when reviewing teammates' PRs to ensure they match the design doc/spec. The reverse of bill-feature-implement. Use when user mentions verify PR, check PR against spec, review against design doc, or verify implementation.
Use when running a governed editorial assignment desk from Readian recommendations through candidate selection and source-backed story packs.
Stress-test a plan or design by walking every branch of the decision tree until reaching shared understanding. Use when user wants to challenge a design, stress-test a plan, get grilled, poke holes, or says "grill me".
Use when reviewing unit tests in a file, current changes, or a commit to flag low-value, tautological, or coverage-only tests that do not validate real behavior. Use when user mentions check test quality, review tests, tautological tests, weak tests, or coverage-padding.
Use when reviewing or building KMP UI surfaces. Today this skill is implemented with Jetpack Compose-specific guidance, but it is the canonical KMP UI review capability so future platform UI guidance can live behind the same slash command. Enforces state hoisting, proper recomposition handling, slot-based APIs, accessibility, theming, string resources, preview annotations, and official UI framework guidelines. Use when user mentions Compose review, UI review, recomposition, state hoisting, or Composable code.
Use when doing end-to-end feature implementation from design doc to verified code. Automatically scales ceremony based on feature size - lightweight for small changes, full orchestration for medium features, and explicit decomposition into resumable subtask specs when work is too large for one reliable execution. Runs each heavy phase (pre-planning, planning, implementation, completeness audit, quality check, PR description) inside its own subagent with a rich self-contained briefing, to keep the orchestrator context small. Code review stays in the orchestrator because it already spawns specialist subagents internally. Use when user mentions implement feature, build feature, implement spec, or feature from design doc.
Use when conducting a thorough Kotlin PR code review across shared, backend/server, or generic Kotlin code, or when providing the baseline Kotlin review layer for Android/KMP reviews. Select shared Kotlin specialists for architecture, correctness, security, performance, and testing, and add backend-focused specialists for API contracts, persistence, and reliability when server signals are present. Produces a structured review with risk register and prioritized action items. Use when user mentions Kotlin review, review Kotlin PR, Kotlin code review, or asks to review .kt files.
Use when conducting a thorough Android/KMP PR code review. Preserve mobile review depth by running the appropriate Kotlin baseline review layer first, then add Android/KMP-specific specialists such as UI and UX/accessibility. Produces a structured review with risk register and prioritized action items. Use when user mentions Android review, KMP review, mobile review, or asks to review Android/KMP changes.
Use when removing an existing skill or platform skill set and cleaning up agent installs, manifests, and supporting links.
Use when recording architectural or implementation decisions in a module/package/area agent/decisions.md file. Use when user mentions record decision, boundary decision, why we chose, decision log, or remember this decision.
Use when updating module/package/area agent/history.md files with reusable, high-signal feature history entries and history hygiene rules. Use when user mentions update history, write history entry, boundary history, or record feature history.
Use when generating a PR title, description, and QA steps from the current branch changes. Works standalone or as part of bill-feature-implement. Use when user mentions write PR description, generate PR, PR title, or create pull request.
Use when you want a generic quality-check entry point that detects the dominant stack in scope and delegates to the matching stack-specific quality-check skill. Use when user mentions run checks, validate, lint, format, quality check, or run quality.
Use when you want a generic code-review entry point that detects the dominant stack in scope and delegates to the matching stack-specific review skill. Use when user mentions code review, review my changes, review this PR, review staged changes, or asks to review code.
Remove feature flags and legacy code after a feature is fully rolled out. Use when a feature flag has been enabled for all users and the legacy path is no longer needed. Safely deletes Legacy classes, removes flag checks, and inlines the winning code path. Use when user mentions remove feature flag, cleanup flag, delete legacy code, or flag fully rolled out.
Enable feature flag mode - all code changes will be guarded by feature flags for safe rollback. Use when implementing new features that need gradual rollout, A/B testing, or safe rollback capability. Applies the Legacy pattern for large changes, factory/DI switching for medium changes, and simple conditionals for small changes. Use when user mentions feature flag, feature toggle, gradual rollout, safe rollback, or guard with flag.
Use when reviewing secrets handling, auth/session safety, sensitive data exposure, and transport/storage security in Kotlin code. Use when user mentions secrets, auth tokens, encryption, sensitive data, or security review in Kotlin code.
Use when reviewing Kotlin backend/server API boundaries including request validation, serialization, HTTP or RPC contracts, status-code mapping, and backward compatibility. Use when user mentions API contract, request validation, response serialization, status codes, or backward compatibility in Kotlin backend.
Use when reviewing architecture, boundaries, DI scopes, and source-of-truth consistency in Kotlin code. Use when user mentions Kotlin architecture, DI scope, module boundaries, or dependency direction in Kotlin code.
Use when reviewing lifecycle, coroutine, threading, and logic correctness risks in Kotlin code. Use when user mentions coroutine leak, race condition, dispatcher misuse, lifecycle-unsafe collection, or threading bug in Kotlin code.
Use when reviewing Kotlin backend/server persistence risks including transaction boundaries, query correctness, migration safety, concurrency, and data-consistency behavior. Use when user mentions database review, transaction boundaries, migration safety, ORM mapping, or query correctness in Kotlin backend.
Use when reviewing UX correctness and accessibility risks, delegating UI-framework-heavy checks to bill-kmp-code-review-ui. Use when user mentions UX review, accessibility, content description, screen reader, or localization review.
Use when reviewing performance risks in Kotlin code, including hot-path work, blocking I/O, latency regressions, and resource waste. Use when user mentions performance, blocking I/O, hot path, memory leak, or latency in Kotlin code.
Use when reviewing Kotlin backend/server reliability risks including timeouts, retries, background work, concurrency under load, caching, and observability-critical failures. Use when user mentions timeout, retry logic, circuit breaker, caching, or observability in Kotlin backend.
Use when reviewing test coverage quality, real test value, regression protection, and test reliability risks in Kotlin code. Use when user mentions test quality, test coverage, mock setup, or test reliability in Kotlin code.
Use when scaffolding a new skill or platform skill set and syncing it to all detected local AI agents (Claude, Copilot, Codex, Opencode). Use when user mentions scaffold skill, create skill set, create skill, new skill, add skill, or sync skill to agents.
Fixture content missing a required H2 section.
Fixture baseline content.
Fixture baseline content for a pack that also declares a quality-check file.
Fixture baseline used to reject non-list declared_code_review_areas.
Fixture baseline pack content used for shell contract coverage.
Fixture quality-check content.
Fixture quality-check content with all five required sections.
Use when creating a new skill and syncing it to all detected local AI agents (Claude, Copilot, GLM, Codex, Opencode). Use when user mentions create skill, new skill, add skill, or sync skill to agents.
Use when conducting a thorough Go PR code review across backend/service projects. Classify changed areas conservatively, select the right specialist review passes for the diff, including real test-value review when tests change. Produces a structured review with risk register and prioritized action items. Use when user mentions Go review, review Go PR, Go code review, or asks to review .go files.
Use when reviewing Kotlin backend/server API boundaries including request validation, serialization, HTTP or RPC contracts, status-code mapping, and backward compatibility. Use when user mentions API contract, request validation, response serialization, status codes, or backward compatibility in Kotlin backend.
Use when reviewing Kotlin backend/server persistence risks including transaction boundaries, query correctness, migration safety, concurrency, and data-consistency behavior. Use when user mentions database review, transaction boundaries, migration safety, ORM mapping, or query correctness in Kotlin backend.
Use when reviewing Kotlin backend/server reliability risks including timeouts, retries, background work, concurrency under load, caching, and observability-critical failures. Use when user mentions timeout, retry logic, circuit breaker, caching, or observability in Kotlin backend.
Use when recording architectural or implementation decisions in a module/package/area agent/decisions.md file. Use when user mentions record decision, boundary decision, why we chose, decision log, or remember this decision.
Enable feature flag mode - all code changes will be guarded by feature flags for safe rollback. Use when implementing new features that need gradual rollout, A/B testing, or safe rollback capability. Applies the Legacy pattern for large changes, factory/DI switching for medium changes, and simple conditionals for small changes. Use when user mentions feature flag, feature toggle, gradual rollout, safe rollback, or guard with flag.
Fixture quality-check content missing the Fix Strategy section.
Fixture baseline used to reject unapproved entries in declared_code_review_areas.
Fixture baseline content.
Fixture file with a required H2 buried inside a fenced code block to prove the loader refuses to accept that as a real heading.
Run the repository's canonical validation commands for governed skill and agent-configuration repositories, then fix issues in the current unit of work without using suppressions. Use when validating skill contracts, routing docs, installers, catalogs, and repo-native validator scripts. Use when user mentions validate skills, validate agent config, run skill validation, or check skill contracts.
Use when conducting a thorough Kotlin backend/server PR code review. Preserve backend review depth by running bill-kotlin-code-review as the baseline Kotlin review layer, then add backend-specific specialists such as API contracts, persistence, and reliability. Produces a structured review with risk register and prioritized action items. Use when user mentions backend Kotlin review, server review, Ktor review, Spring review, or backend PR review.
Use when you want a generic code-review entry point that detects the dominant stack in scope and delegates to the matching stack-specific review skill. Use when user mentions code review, review my changes, review this PR, review staged changes, or asks to review code.
Remove feature flags and legacy code after a feature is fully rolled out. Use when a feature flag has been enabled for all users and the legacy path is no longer needed. Safely deletes Legacy classes, removes flag checks, and inlines the winning code path. Use when user mentions remove feature flag, cleanup flag, delete legacy code, or flag fully rolled out.
Experimental subagent-based variant of bill-feature-implement. Runs each heavy phase (pre-planning, planning, implementation, completeness audit, quality check, PR description) inside its own subagent with a rich self-contained briefing, to keep the orchestrator context small. Code review stays in the orchestrator because it already spawns specialist subagents internally. Use when you want the same end-to-end feature workflow but with less orchestrator context pollution.
Use when generating a PR title, description, and QA steps from the current branch changes. Works standalone or as part of bill-feature-implement. Use when user mentions write PR description, generate PR, PR title, or create pull request.
Use when you want a generic quality-check entry point that detects the dominant stack in scope and delegates to the matching stack-specific quality-check skill. Use when user mentions run checks, validate, lint, format, quality check, or run quality.
Use when reviewing architecture, boundaries, package design, and source-of-truth consistency in Go backend or service changes. Use when user mentions Go architecture, package design, cmd/ layout, internal/ boundaries, or dependency direction in Go.
Use when reviewing performance risks in Go backend/service code, including hot-path work, blocking I/O, query-shape issues, inefficient DB/network access, marshaling overhead, buffering, goroutine churn, and resource waste. Use when user mentions performance, N+1, allocation churn, goroutine storm, or marshaling overhead in Go.
Use when reviewing Go backend/service reliability risks including timeouts, retries, background work, concurrency under load, caching, and observability-critical failures. Use when user mentions timeout, retry, shutdown, health check, backpressure, or rate limiting in Go.
Use when reviewing security risks in changed Go backend/service code including auth/session safety, secrets handling, trust boundaries, sensitive data exposure, injection, file handling, and output encoding. Use when user mentions auth, secrets, TLS, injection, SSRF, or security review in Go.
Use when reviewing architecture, boundaries, DI scopes, and source-of-truth consistency in Kotlin code. Use when user mentions Kotlin architecture, DI scope, module boundaries, or dependency direction in Kotlin code.
Use when reviewing secrets handling, auth/session safety, sensitive data exposure, and transport/storage security in Kotlin code. Use when user mentions secrets, auth tokens, encryption, sensitive data, or security review in Kotlin code.
Use when reviewing test coverage quality, real test value, regression protection, and test reliability risks in Kotlin code. Use when user mentions test quality, test coverage, mock setup, or test reliability in Kotlin code.
Use when conducting a thorough Kotlin PR code review across shared or generic Kotlin code, or when providing the baseline Kotlin review layer for Android/KMP and backend/server reviews. Select shared Kotlin specialists for architecture, correctness, security, performance, and testing. Produces a structured review with risk register and prioritized action items. Use when user mentions Kotlin review, review Kotlin PR, Kotlin code review, or asks to review .kt files.
Use when reviewing PHP backend/server persistence risks including transaction boundaries, query correctness, migration safety, concurrency, and data-consistency behavior. Use when user mentions database review, Eloquent, query builder, migrations, transactions, or ORM in PHP.
Use when reviewing behavior correctness, edge cases, retry and idempotency behavior, ordering bugs, and concurrency-sensitive logic in PHP backend/server changes. Use when user mentions PHP correctness, state transitions, retry logic, idempotency, or ordering bugs in PHP.
Use when reviewing security risks in changed PHP/backend code including auth/session safety, secrets handling, trust boundaries, sensitive data exposure, injection, file handling, and output encoding. Use when user mentions auth, XSS, CSRF, injection, upload safety, or security review in PHP.
Use when conducting a thorough PHP PR code review across backend/server projects. Classify changed areas conservatively, select the right specialist review passes for the diff, including real test-value review when tests change. Produces a structured review with risk register and prioritized action items. Use when user mentions PHP review, review PHP PR, PHP code review, or asks to review .php files.
Use when updating module/package/area agent/history.md files with reusable, high-signal feature history entries and history hygiene rules. Use when user mentions update history, write history entry, boundary history, or record feature history.
Use when reviewing or building KMP UI surfaces. Today this skill is implemented with Jetpack Compose-specific guidance, but it is the canonical KMP UI review capability so future platform UI guidance can live behind the same slash command. Enforces state hoisting, proper recomposition handling, slot-based APIs, accessibility, theming, string resources, preview annotations, and official UI framework guidelines. Use when user mentions Compose review, UI review, recomposition, state hoisting, or Composable code.
Stress-test a plan or design by walking every branch of the decision tree until reaching shared understanding. Use when user wants to challenge a design, stress-test a plan, get grilled, poke holes, or says "grill me".
Use when conducting a thorough code review for governed skill, prompt, and agent-configuration repositories. Focus on routing correctness, contract drift, installer safety, portability, and docs/tests/catalog consistency. Produces a structured review with risk register and prioritized action items. Use when user mentions review skill config, review agent config, routing review, installer review, or skill repository review.
Use when doing end-to-end feature implementation from design doc to verified code. Automatically scales ceremony based on feature size — lightweight for small changes, full orchestration for large features. Runs each heavy phase (pre-planning, planning, implementation, completeness audit, quality check, PR description) inside its own subagent with a rich self-contained briefing, to keep the orchestrator context small. Code review stays in the orchestrator because it already spawns specialist subagents internally. Use when user mentions implement feature, build feature, implement spec, or feature from design doc.
Verify a PR against a task spec — extract acceptance criteria, decide whether rollout expectations matter, audit feature-flag behavior only when the spec or diff requires it, run full code review, and audit completeness. Use when reviewing teammates' PRs to ensure they match the design doc/spec. The reverse of bill-feature-implement. Use when user mentions verify PR, check PR against spec, review against design doc, or verify implementation.
Use when reviewing unit tests in a file, current changes, or a commit to flag low-value, tautological, or coverage-only tests that do not validate real behavior. Use when user mentions check test quality, review tests, tautological tests, weak tests, or coverage-padding.
Use when reviewing Go backend/service API boundaries including request validation, serialization, HTTP or RPC contracts, status-code mapping, and backward compatibility. Use when user mentions API contract, HTTP handler, gRPC, struct tags, omitempty, or status codes in Go.
Use when reviewing Go backend/service persistence risks including transaction boundaries, query correctness, migration safety, concurrency, and data-consistency behavior. Use when user mentions database review, SQL, transaction, migration, GORM, sqlx, or query correctness in Go.
Use when reviewing behavior correctness, edge cases, goroutine/channel safety, context cancellation, error handling, and concurrency-sensitive logic in Go backend/service changes. Use when user mentions goroutine leak, channel safety, context cancellation, race condition, or nil handling in Go.
Use when reviewing test coverage quality, real test value, regression protection, and test reliability risks in Go backend/service code. Use when user mentions test quality, t.Run, t.Parallel, fuzz test, flaky test, or test coverage in Go.
Run the project's canonical quality-check command(s) and systematically fix all issues without using suppressions. Use when running repo-defined checks, tests, linting, static analysis, formatting, or security validation in Go projects. Fixes issues properly at the root cause instead of suppressing them. Use when user mentions run Go checks, go vet, golangci-lint, go test, or fix Go lint issues.
Use when reviewing UX correctness and accessibility risks, delegating UI-framework-heavy checks to bill-kmp-code-review-ui. Use when user mentions UX review, accessibility, content description, screen reader, or localization review.
Use when conducting a thorough Android/KMP PR code review. Preserve mobile review depth by running the appropriate Kotlin baseline review layer first, then add Android/KMP-specific specialists such as UI and UX/accessibility. Produces a structured review with risk register and prioritized action items. Use when user mentions Android review, KMP review, mobile review, or asks to review Android/KMP changes.
Use when reviewing performance risks in Kotlin code, including hot-path work, blocking I/O, latency regressions, and resource waste. Use when user mentions performance, blocking I/O, hot path, memory leak, or latency in Kotlin code.
Use when reviewing lifecycle, coroutine, threading, and logic correctness risks in Kotlin code. Use when user mentions coroutine leak, race condition, dispatcher misuse, lifecycle-unsafe collection, or threading bug in Kotlin code.
Run ./gradlew check and systematically fix all issues without using suppressions. Use when running Gradle checks, fixing lint errors, formatting issues, test failures, or deprecation warnings in Gradle/Kotlin projects. Fixes issues properly at the root cause instead of suppressing them. Use when user mentions gradlew check, Kotlin lint, ktfmt, detekt, or fix Gradle warnings.
Use when reviewing PHP backend/server API boundaries including request validation, serialization, HTTP or RPC contracts, status-code mapping, and backward compatibility. Use when user mentions API contract, Laravel routes, controllers, request validation, or response serialization in PHP.
Use when reviewing architecture, boundaries, layering, and source-of-truth consistency in PHP backend or server-rendered changes. Use when user mentions PHP architecture, layering, ports/adapters, CQRS, module boundaries, or DDD in PHP.
Use when reviewing performance risks in PHP backend/server code, including hot-path work, blocking I/O, ORM/query-shape issues, inefficient DB/network access, serialization/rendering overhead, buffering, and resource waste. Use when user mentions N+1, query performance, rendering waste, serialization overhead, or hot path in PHP.
Use when reviewing PHP backend/server reliability risks including timeouts, retries, background work, concurrency under load, caching, and observability-critical failures. Use when user mentions job reliability, queue retry, cache invalidation, external client timeout, or observability in PHP.
Use when reviewing test coverage quality, real test value, regression protection, and test reliability risks in PHP backend/server code. Use when user mentions test quality, PHPUnit, contract tests, test coverage, or test reliability in PHP.
Run the project's canonical quality-check command(s) and systematically fix all issues without using suppressions. Use when running repo-defined checks, tests, linting, static analysis, architecture checks, or security validation in PHP/Laravel projects. Fixes issues properly at the root cause instead of suppressing them. Use when user mentions run PHP checks, PHPStan, PHP CS Fixer, Laravel Pint, or fix PHP lint issues.
Run ./gradlew check and systematically fix all issues without using suppressions. Use when running Gradle checks, fixing lint errors, formatting issues, test failures, or deprecation warnings in Gradle/Kotlin projects. Fixes issues properly at the root cause instead of suppressing them. Use when user mentions gradlew check, Kotlin lint, ktfmt, detekt, or fix Gradle warnings.
Fixture content with all required sections but bad contract version.
Fixture baseline content.
Fixture baseline to exercise rejection of extra declared_files.areas entries.
Fixture skill.