agentleFS
garbageekgarbageek/claude-plugins/skills/audit-triage/SKILL.md

audit-triage

Skill1 starsChanged 35 days agoMITView on GitHub

Use when existing audit tickets need prioritization, dependency ordering, shared-file or shared-cause fix groups, or a roadmap. Do not create tickets or edit implementation code.

The file

We split this skill into the rules it gives an agent

12 rules. About 0 are common, 6 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read the runtime invocation contract first.

    line 9
  2. Distinctive:

    If the workspace is absent or no tickets exist, report that state without initializing it.

    line 24
  3. Very specific:

    Use deps add/remove for genuine prerequisites.

    line 63
  4. Very specific:

    Check next --for resolution --json: ticket is the selection.

    line 68
  5. Distinctive:

    read tickets and canonical export.

    line 74
  6. Distinctive:

    add or remove dependency edges.

    line 76
  7. Very specific:

    write optional roadmap files under audit/triage/.

    line 77
  8. Distinctive:

    do not create tickets unless the user explicitly switches to discovery.

    line 81
  9. Distinctive:

    do not edit implementation code.

    line 82
  10. Distinctive:

    do not resolve or verify tickets.

    line 83
  11. Very specific:

    do not mark tickets PASS.

    line 84
  12. Very specific:

    Stop when doctor accepts the metadata and next --for resolution --json returns the expected highest-priority unblocked ticket and coherent …

    line 88

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “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 Triage

Read the runtime invocation contract first. Use the bundled Python CLI after resolving AUDIT_PLUGIN_ROOT and AUDIT_PROJECT_DIR as described there. A separately configured audit_* MCP server is an optional equivalent with an explicit absolute root.

Purpose: plan coherent fixes across tickets, not just assign statuses. Resolve dependencies and shared-change opportunities, then write decisions, scores, phases, and fix groups to ticket headers. A prose roadmap is optional; next --for resolution reads only ticket state.

First move from empty context

python3 "${AUDIT_PLUGIN_ROOT}/scripts/audit.py" --root "${AUDIT_PROJECT_DIR}" doctor
python3 "${AUDIT_PLUGIN_ROOT}/scripts/audit.py" --root "${AUDIT_PROJECT_DIR}" export --json
python3 "${AUDIT_PLUGIN_ROOT}/scripts/audit.py" --root "${AUDIT_PROJECT_DIR}" summary

If the workspace is absent or no tickets exist, report that state without initializing it. Do not create findings during triage. Treat dependency-wait warnings separately from structural errors; a pending prerequisite is not a corrupt audit.

Update executable metadata

python3 "${AUDIT_PLUGIN_ROOT}/scripts/audit.py" --root "${AUDIT_PROJECT_DIR}" triage set 001 --impact 5 --effort 2 --p-level P0 --decision FIX --phase critical-path
python3 "${AUDIT_PLUGIN_ROOT}/scripts/audit.py" --root "${AUDIT_PROJECT_DIR}" deps add 001 --depends-on 003
python3 "${AUDIT_PLUGIN_ROOT}/scripts/audit.py" --root "${AUDIT_PROJECT_DIR}" triage set 001 --fix-group shared-parser-fix

Dependency semantics:

003 blocks 001

Ticket 001 is skipped by python3 "${AUDIT_PLUGIN_ROOT}/scripts/audit.py" --root "${AUDIT_PROJECT_DIR}" next --for resolution until 003 reaches PASS.

Scoring

FieldValuesMeaning
Impact1-5User/data/release impact
Effort1-5Implementation effort
Priority1-25Impact × (6 - Effort)
P-levelP0-P3Execution priority
DecisionFIX/DEFER/WONTFIX/DUPLICATEOperational decision
Phasecritical-path/quick-wins/feature-parity/backlogExecution group
Fix GroupOptional short slugTickets that can share one coherent code change

Grouping and scheduling

  • Inspect affected files, call paths, overlapping changes and underlying causes before choosing groups. Prefer one coherent edit session for compatible issues in the same source file or a shared component, rather than repeating the same investigation and patch multiple times.
  • Assign the same --fix-group only when work can actually be resolved together. Shared files alone are a suggestion, not proof of a shared fix; identify conflicting edits and dependencies before grouping.
  • Use deps add/remove for genuine prerequisites. A deferred, rejected or blocked ticket must not become executable merely because its file matches a selected group. Only PASS dependencies unblock resolution.
  • Apply every triage decision through the CLI before producing a report or roadmap. The report is derived from export --json and never feeds the queue.
  • Check next --for resolution --json: ticket is the selection; related_tickets are additional eligible FIX suggestions for shared work, not automatically authorized extra tickets.

Allowed actions

  • read tickets and canonical export;
  • set triage metadata;
  • add or remove dependency edges;
  • write optional roadmap files under audit/triage/.

Forbidden actions

  • do not create tickets unless the user explicitly switches to discovery;
  • do not edit implementation code;
  • do not resolve or verify tickets;
  • do not mark tickets PASS.

Stop condition

Stop when doctor accepts the metadata and next --for resolution --json returns the expected highest-priority unblocked ticket and coherent suggestions.

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.