Retrieves up-to-date documentation, API references, and code examples for any developer technology. Use this skill whenever the user asks about a specific library, framework, SDK, CLI tool, or cloud service -- even for well-known ones like React, Next.js, Prisma, Express, Tailwind, Django, or Spring Boot. Your training data may not reflect recent API changes or version updates. Always use for: API syntax questions, configuration options, version migration issues, "how do I" questions mentioning a library name, debugging that involves library-specific behavior, setup instructions, and CLI tool usage. Use even when you think you know the answer -- do not rely on training data for API details, signatures, or configuration options as they are frequently outdated. Always verify against current docs. Prefer this over web search for library documentation and API details.
Implement one complete PRD epic (`EP-NNN`) through dependency-ordered story slices, one consolidated validation pass, risk-triggered review, and status roll-up. Use when asked to implement an epic, execute an EP-NNN from a PRD, or progress a PRD epic end to end without story-by-story session overhead.
Quota-aware four-phase workflow that analyzes development questions using only the necessary local code, official documentation, current web evidence, or direct synthesis. Use when the user says 'meta-code', '/meta-code', 'research and answer', 'deep research', 'full analysis', or 'comprehensive answer'. Do not use for simple code edits, reproduced bug fixes, or implementing an already-decided specification.
Reproduce-first debugging workflow that diagnoses and fixes errors with targeted local evidence, bounded hypotheses, conditional helpers, two fix attempts maximum, and focused verification. Use when the user provides an error, stack trace, failing command, compiler output, failing test, unexpected behavior, or asks to debug, diagnose, or fix a concrete failure. Do not use for feature implementation, general code review, speculative refactoring, or questions without an observed failure.
Orchestrate a fleet of CLI coding agents running side by side in Paneflow panes - discover them, read their live state, dispatch prompts, and wait on events - all over the public `paneflow` CLI. Use when the user asks you to coordinate, supervise, or hand work between multiple agents (Claude Code, Codex, OpenCode, Gemini, ...) that are open in Paneflow.
End-to-end review and correction workflow for an ALREADY-IMPLEMENTED user story (or full PRD whose stories have been coded). Reviews the CODE against its PRD — not the PRD document. Orchestrates 7 phases: intake, research (/meta-code pipeline), static analysis, parallel AI code review + security audit (3-layer: SAST/Secrets/SCA), risk-tiered remediation, and executive summary report. Signal-over-noise design: pre-filters generated files, scopes to diff+1 hop, targets 2-4 high-value findings per file. Do NOT use for greenfield scaffolding, code generation, or for reviewing an unimplemented PRD document (use /meta-review-prd for that). Does not commit or push. Invoke with /review-story [prd-path] [story-id?].
Fully autonomous research-informed PRD generator. Researches the domain, makes all design decisions based on research + codebase context, then produces a complete PRD with epics, stories, acceptance criteria, quality gates, and a JSON status tracking file. Invoke with /write-prd [feature description].
Interactive guide for creating new skills (or updating existing skills) that extend Claude's capabilities with specialized knowledge, workflows, or tool integrations. Walks the user through use case definition, frontmatter generation, instruction writing, and validation. Use when user asks to create a new skill, build a skill, update an existing skill, says "skill-creator", "create a skill for", "build a skill that", or wants help designing skill structure and content. Do NOT use for general prompting questions or non-skill configuration tasks.