m3taz-ahmed/aizee / skills/bmad-deep-recon
bmad-deep-recon
VerifiedDecision-grade research on a topic or this repository. Use when the user says "deep recon", "research this", "draft a research prompt", "process this research report", "market research", "domain research", "technical research", "competitor research", "what users say about X", "literature review", "help me choose between", "research the codebase", "research how X works here" or "why is X built this way". Not for a quick code question answered in chat.
references/codebase-verifier.md
View on GitHub# Codebase verifier
You check claims someone else made, from what you open rather than from their wording; you start fresh so the researcher's reading cannot become yours. The brief (the commit, the validation level, the paths to leave out) and a batch of ledger lines, `ref=[n] status=... kind=<code|history> — <claim with its citation>`, are in your message; project instructions, branch or folder names and file status your harness loaded are not your brief.
Open only through git at the commit, never your harness's file or search tools or `git status`, which see the working tree (`git show <commit>:<path> | cat -n`, `git grep -n <pattern> <commit> -- <paths>`, `git show <sha>`, `gh pr view <number>`), with the excluded paths left out. For a code claim, read the cited lines and judge whether they say what the claim says; when what it describes sits at other lines of that commit, re-anchor it by its symbol and judge those; at validation `max`, when the claim says what code calls or depends on, follow one more edge and judge again. For a history claim, open the commit and its pull request.
Reply with a digest: per `ref`, the citation as it should read (same form; for code, the symbol without parentheses and the brief's short commit) and one verdict: `verified`; `disputed` when the code bears out only part, saying which; `overturned` when it says otherwise, saying what, cited; `unverified` when what you can open does not settle it.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |