agentleFS
alegaussalegauss/claude-tray/.claude/skills/audit/SKILL.md

audit

Skill7 starsChanged 12 days agoApache-2.0View on GitHub

Full audit of Claude Code Tray — build, self-test, parallel per-partition scanners, adversarial verification, and reconciliation of the roadkeep backlog (close done tasks, retire outdated ones, reword unclear ones, file confirmed findings). Use when asked to audit the repo, hunt for bugs across the codebase, or reconcile the roadmap against the code.

193 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use delight, teach-impeccable and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • HK-hubSkill

    delight

    Add moments of joy, personality, and unexpected touches that make interfaces memorable and enjoyable to use. Elevates functional to delightful. Use when the user asks to add polish, personality, animations, micro-interactions, delight, or make an interface feel fun or memorable.

    Also in 31 projects that use this

  • HK-hubSkill

    teach-impeccable

    One-time setup that gathers design context for your project and saves it to your AI config file. Run once to establish persistent design guidelines.

    Also in 18 projects that use this

  • zwy123zwySkill

    normalize

    将用户随意输入的提示词规范化为结构清晰、便于大模型执行的标准化提示词。适用于需要整理、统一提示词格式时调用。

    Also in 18 projects that use this

  • ww-w-aiSkill

    pm-discovery

    PM Agent Team - Automated product discovery, strategy, and PRD generation. Runs PM analysis workflow to produce comprehensive PRD before PDCA Plan phase.

    Also in 3 projects that use this

  • ww-w-aiSkill

    qa-phase

    Run QA tests at various levels (L1-L5) for features under PDCA management. Integrates with MCP tools to execute structured test plans.

    Also in 3 projects that use this

  • +7 more from HK-hub

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 1 is distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 173 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run the steps in order. Do not skip a step.

    line 8
  2. Very specific:

    Every failure here is a finding (path:line | … | … | the command).

    line 19
  3. Very specific:

    Use the Agent tool with subagent_type: "scanner", prompt Partition: <paths>:

    line 24
  4. Very specific:

    Never hand-edit ROADMAP.md, CHANGELOG.md, IMPROVEMENTS.md or STRATEGY.md.

    line 54
  5. Very specific:

    Do not commit — the user decides (one task per commit, via run-commit.cmd -m).

    line 68

“Common” is measured: kept by at least 6 in 10 of the 173 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

/audit

Run the steps in order. Do not skip a step; if one cannot run, say so in the report.

1. Baseline checks (sequential — never two builds at once)

dotnet build -c Debug --nologo -warnaserror
powershell -NoProfile -Command "$p = Start-Process 'bin\Debug\net10.0-windows\win-x64\ClaudeTray.exe' -ArgumentList '--selftest','--quick' -Wait -PassThru -RedirectStandardOutput $env:TEMP\st.log; Get-Content $env:TEMP\st.log; exit $p.ExitCode"
cd site && npm run typecheck && npm test
roadkeep lint

Every failure here is a finding (path:line | … | … | the command). UI cases (dotnet test tests\ClaudeTray.Cases) are NOT run on this desk — they run in the VM.

2. Scan — one scanner per partition, all in ONE message (parallel)

Use the Agent tool with subagent_type: "scanner", prompt Partition: <paths>:

#Partition
P1src/Cli/
P2src/Ui/
P3src/Usage/
P4src/Tray/, src/Core/, src/Profiles/
P5src/Context/, tests/ClaudeTray.Cases/, cases/
P6site/src/, site/scripts/, build/, scripts/, lang/, .github/workflows/

3. Deduplicate

Merge the lines from step 1 and 2. Same path:line, or same issue at the same call site → one line. Drop NO FINDINGS.

4. Verify findings

One verifier agent, prompt MODE: FINDINGS + the full deduplicated list. Keep only CONFIRMED.

5. Verify the backlog

roadkeep list --json

One verifier agent, prompt MODE: ROADMAP + that task list + the CONFIRMED findings.

6. Reconcile — through roadkeep only, never by editing the files

Never hand-edit ROADMAP.md, CHANGELOG.md, IMPROVEMENTS.md or STRATEGY.md.

  • ALREADY DONE → roadkeep ship <id> --why "<what now works, one sentence.>"
  • OUTDATED → roadkeep retire <id> --reason "<why it no longer applies.>" (use --folds-into <id> when another open line absorbs it)
  • NEEDS REWORDING → roadkeep restate <id> --symptom "<text>" and/or roadkeep amend <id> --why "<text>"
  • NEW (CONFIRMED and not covered) → first roadkeep budget --symptom "<text>", then roadkeep add --block <L> --ref <anchor> --symptom "<text>" --why "<text>" --section "<title>" --section-body-file <scratch file>. --ref is mandatory here (ref_scheme = "outline"); find a free anchor with roadkeep anchors. Pass long text via a variable or file, never inline in PowerShell. Reuse an existing block.
  • STILL VALID → no action.

Then roadkeep lint; if it reports anything, roadkeep repair, then roadkeep lint again. Do not commit — the user decides (one task per commit, via run-commit.cmd -m).

7. Report — counts only

Findings: <n> confirmed, <n> false positive, <n> unverifiable
Tasks:    <n> kept, <n> closed, <n> removed, <n> rewritten, <n> added

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.