Loading...
Loading...
Found 50 Skills
Write a commit message body by deriving it from the code, not by summarizing the diff. Weighs the change, then either answers seven questions against the implementation — what forces it, which call paths reach it, how it failed, what breaks on update, what the tests pin — or produces a two-to-three-line body for mechanical and generated changes. Not for full commit creation.
Contribute changes to the Feynman repository itself. Use when the task is to add features, fix bugs, update prompts or skills, change install or release behavior, improve docs, or prepare a focused PR against this repo.
Lightweight workflow for straightforward changes — plan → implement → optional PR. Direct-commit by default; synthesize is opt-in via synthesisPolicy or a runtime request_synthesize event. Use for trivial fixes, config tweaks, single-file changes, or exploratory work that doesn't warrant subagent dispatch or two-stage review. Triggers: 'oneshot', 'quick fix', 'small change', or oneshot.
Use when starting a new feature, fix, or refactor that needs explicit planning before code changes. Reads .handoff/config.md (and backlog if present) and writes a structured plan.md to .handoff/. Supports multi-phase plans for large work, and risk tags on change list items. Does NOT modify code or run build commands. Pair with /execute and /verify. Part of the agent-handoff bundle (4 skills) — install /setup-handoff, /plan, /execute, /verify together.
Multi-AI Cross-Review Workflow. Conduct cross-reviews of code changes between different AI tools (Claude, Codex, Gemini, etc.). Trigger scenarios: When users call /discuss-multi-ai, mention asking other AIs to review changes, or post a review request containing a commit hash. This skill should also be triggered even if users simply say "Help me generate review text for another AI" or "Send this change to other AIs for review."
Produces a single-story walkthrough of AI-authored code changes from runtime trigger to final behavior, weaving changed and unchanged code into one narrative with annotated diffs, trade-offs, alternatives, and risk analysis. Use when asked to "explain what changed", "walk me through this diff", "summarize agent edits", "show how this feature works", or "explain this implementation step by step".
El orquestador ("el jefe") de la software factory. Recibe un cambio — feature nuevo, bug, ajuste de UI, cambio de datos/schema, hardening, o trivial — lo CLASIFICA, elige la ruta mínima de subagentes/skills, los corre (en paralelo si los territorios son disjuntos) con gate humano y verificación, y lo lleva hasta el PR. Use cuando traés un cambio para que la fábrica lo procese, o al arrancar cualquier trabajo no-trivial que convenga rutear bien. No es un pipeline fijo: rutea según el tipo de cambio.
Refresh AI's understanding of code. Use this skill when the user mentions terms like "refresh", "re-understand", "refresh cache", "reload", etc. This skill compares all modified files, re-reads and understands the code that may have been modified by humans, ensuring that the AI's context is synchronized with the latest code state.
Use when the user wants to verify their understanding of a branch's code changes by being quizzed on runtime behavior, assumptions, failure points, and edge cases instead of just reading diffs
Analyse code changes since the last docs update and refresh the project's documentation files. Use when code has changed and documentation needs to be updated, after implementing new features, or before a milestone commit.
Implement narrowly scoped code changes with focused validation. Use for small bug fixes, tweaks, and features.
Find downstream breakage a code change could cause beyond the files it directly touches.