菜单
Diagnose and fix a bug using the SEB method (Screenshot, Error message, Before/After) plus the bisect trick for unknown regressions, gathering real evidence before guessing at a fix. This is the fuller version of /fix-bug — use directly for trickier bugs or when the quick pass doesn't find the cause.
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
| name | debug-seb |
| description | Diagnose and fix a bug using the SEB method (Screenshot, Error message, Before/After) plus the bisect trick for unknown regressions, gathering real evidence before guessing at a fix. This is the fuller version of /fix-bug — use directly for trickier bugs or when the quick pass doesn't find the cause. |
The founder can't read a stack trace, but can take a screenshot and copy error text. The job here is to force evidence collection before guessing, with a deterministic fallback for when there's no clear "before."
Collect S-E-B before touching code — don't skip to "just fix it."
Screenshot: use the playwright MCP server to capture one if the
founder can't easily grab it themselves — it's local browser automation
and explicitly "not a security boundary," so it's fine to use freely.
Error text: copied verbatim, not paraphrased — ask again if it was
paraphrased. Before/after: the last change before it broke, if
known.
If "last working state" is unknown, bisect the recent history. Don't stop after an arbitrary number of commits — search backward, testing after each step, until you isolate the actual first bad commit. A fixed 3-commit rollback can miss a regression introduced further back.
Diagnose in plain terms what went wrong before fixing — one or two sentences, no code-level jargon the founder can't use.
Fix only the isolated cause, not surrounding code, to avoid introducing a second regression while chasing this one.
Verify with /verify-path before declaring it fixed, confirming
the step-1 repro no longer reproduces.
/refactor-safely instead.Run the SHIP checklist (Stable, Hidden secrets, Inputs validated, Performance) immediately before every deploy, using the agent's own tools directly. Use right before any git push to a branch that auto-deploys, after /verify-path and /security-audit have already passed.
Run a short retrospective right after a feature actually ships (post /ship-checklist Green, post-deploy) — capture what worked, what didn't, and any surprise, then feed genuine lessons forward into PRD.md's non-goals or behavior rules. Use once per shipped feature/milestone, not on a calendar cadence (that's /weekly-plan's reconciliation step).
Orient a founder who has never used an AI coding agent (Claude Code, OpenCode, or Codex CLI) before to how the *tool* works, before anything about their product idea comes up. Use this first, before /validate-demand, whenever the founder seems unfamiliar with AI coding agents generally — confused by terminal/CLI concepts, unsure what a "slash command" is, or asking how to talk to the agent at all.
Turn a rough product idea into a structured Founding Prompt using the BRIEF framework, then generate PRD.md, AGENTS.md, and llms.txt from it. Use this once demand is validated (after /validate-demand) and before any prototyping begins.
Reverse-engineer a PRD.md/AGENTS.md from a codebase that already exists (with no PRD, or one from another tool/convention), instead of eliciting from scratch. Use this instead of /validate-demand + /founding-prompt whenever a project already has code before founder-os is installed.
Commit in small, atomic "save points" (roughly one logical change each) during any non-trivial task, so a broken step is one git revert/reset away from a known-good state instead of losing a whole session's progress. Use during any multi-step build/fix/refactor, not just at the end.