srnnkls-tropos

Review code using generic review methodology

Start a red vs. blue team debate on a topic

**The user triggered this hint because focus leaked into your recent output.**

Execute implementation tasks - routes to task-dispatch for specs, or code/docs-implement for ad-hoc work

Review a GitHub PR with inline comments

Archive completed development spec

Resolve markers in validation.yaml interactively

Create spec documents for task

Generate GitHub issue drafts from spec

Promote spec from draft to active stage

Review a spec with multiple AI reviewers

Update spec status by analyzing git history

Implement task requirements following TDD

Review changes, provide actionable feedback

Write tests and verify completeness

Ultra-concise bash command patterns. Use when constructing shell commands or one-liners.

Resolve ambiguities interactively with tracked changes. Works with spec-create, spec-review, code-review, and other skills.

Systematic debugging with root cause tracing. Use when encountering bugs, test failures, or unexpected behavior - find root cause before attempting fixes, trace backward through call chain.

Language-specific coding guidelines. Use when implementing code in Python or other supported languages.

Technical evaluation of code review feedback. Use when receiving review feedback - requires verification before implementing, no performative agreement, push back when technically wrong.

Code review methodology. Use when reviewing code locally or preparing for a PR review.

Enforce test-driven development (RED-GREEN-REFACTOR). Use when implementing features, fixing bugs, or changing behavior - write failing test first, then minimal code to pass.

Start structured red vs. blue team debates via subagents. Use when exploring a topic from multiple adversarial perspectives.

General reference documents by domain. Use when creating architecture docs, strategy documents, specs, or guides.

Manage dotfiles using dotter (symlink manager and templater). Use when deploying, adding, removing, or organizing configuration files in ~/dotfiles.

Create isolated git worktrees with safety verification. Use when starting feature work needing isolation or before executing plans - systematic directory selection and baseline verification.

Test Claude Code hooks in isolation and via integration. Use when developing, debugging, or validating hook behavior.

Review GitHub PRs with inline comments and structured summaries. Use when reviewing PRs via gh CLI.

Create new Claude Code skills following project patterns and best practices. Use when building new skills, extracting reusable capabilities, or converting commands to skills.

Archive completed development specs from ./specs/active/ to ./specs/archive/, updating documents with completion status and maintaining archive index. Use when finishing tasks or moving completed work to archive.

Create spec documents (spec.md, context.md, tasks.yaml, dependencies.yaml, validation.yaml). Receives validation data from spec-validate.

Generate GitHub issue drafts from spec directories, creating initiative/feature/task markdown files with gh CLI commands. Use when converting specs to GitHub issues or setting up issue tracking for features.

Promote spec from draft to active stage. Use after spec review passes or when ready to begin implementation.

Multi-agent spec review with parallel Claude/OpenCode reviewers. Use after spec-create or standalone via /spec.review.

Update spec documents by analyzing git history to sync task status with reality.

Validation loop for speccing. Clarifies requirements through structured questioning before document creation.

Evidence-based completion claims. Use before claiming work is complete, fixed, or passing - requires running verification commands and confirming output before any success claims.

Resume spec implementation from checkpoint. Use when continuing task-dispatch after context limit or session break.

Subagent-driven task execution with TDD workflow. Dispatches tester subagent (writes failing tests) then implementer subagent (makes tests pass), with batch review.