菜单
Run the full Forge verification suite (format:check + lint + typecheck + build + test). Use this before claiming any task complete or before opening a PR.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Deep review — random exploration + cross-agent review, alternating. Finds integration bugs that self-review misses. Run every 30–60 min during active implementation.
Overview of Forge's Agentic Coding Flywheel methodology. Use to route between plan, bead, and code space. See FLYWHEEL.md at the repo root.
End-of-session ritual. Close finished beads, file new ones for remaining work, run verify, commit in logical groupings, push. Work is not complete until git push succeeds.
Start a new Flywheel planning round. Produce initial markdown plan, route to competing frontier models, synthesize hybrid, refine 4–5×. Use before any coding begins on a new feature.
Translate a polished markdown plan into self-contained beads in .beads/beads.jsonl. Every bead must be executable without reopening the plan. Use after the plan has converged.
Polishing pass over all open beads. Run 4–6 times. "Check your beads N times, implement once." Use after plan-to-beads and between refinement rounds.
| name | verify |
| description | Run the full Forge verification suite (format:check + lint + typecheck + build + test). Use this before claiming any task complete or before opening a PR. |
Run the exact check sequence that CI runs, in the order that gives the fastest failure signal. Do not mark a task complete until every step below passes.
npm run format:check
npm run lint
npm run typecheck
npm run build
npm test
format:check fails, run npm run format and re-run the suite.lint fails, fix the lint errors directly — never add
eslint-disable without a comment explaining why.typecheck or build fails, fix the types. No as any casts to
silence the compiler.test fails, read the failing test first, then the code under test.
Never skip, .only, or comment out a test to make the suite green.After the suite passes, print a one-line summary:
verify: format ✓ lint ✓ typecheck ✓ build ✓ test ✓ (N tests)
If anything failed, print which step and the first error. Do not claim success unless every step passed.