agentleFS
BirdizBirdiz/skills/skills/audit/audit-setup/SKILL.md

audit-setup

Skill2 starsChanged 15 days agoMITView on GitHub

Configure the auditor profile once per machine (agent, tools, severity scale, report language). Run before the first audit, or when other audit skills report a missing profile.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

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

    Write the auditor profile: what stays true across engagements.

    line 11

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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-setup

Version: 0.26.0 (audit skills, see CHANGELOG)

Write the auditor profile: what stays true across engagements. Scope, access and client belong to audit-kickoff.

Output: ~/.config/audit-skills/profile.md

Process

  1. Detect: shell; docker (any tool an axis names can then run from a pinned image); which of semgrep, gitleaks, trivy, lizard, scc, jscpd, lighthouse, pa11y, axe, hadolint, zizmor and language-native audit commands (composer, npm, pip-audit...) are on PATH; a reachable code-graph tool (e.g. a GitNexus MCP server); a browser automation tool; an issue tracker's CLI (gh, glab, tea) or MCP server; sub-agents.
  2. Confirm the detections in one message, asking only about ambiguous ones.
  3. Interview in rounds: each round asks the frontier (questions whose prerequisites are settled), at most three, numbered, each with a recommended answer ("1 ok, 2 b, 3 ok"), through the agent's structured question tool when it has one. Two rounds cover:
    • auditor background: technical peer (default) or not;
    • default report language;
    • severity scale: the finding-format default, or a custom one with the same number of levels;
    • frameworks cited by default (OWASP ASVS, WCAG 2.2 or RGAA, CNIL guidance...);
    • where engagement workspaces live.
  4. Write the profile:
# Auditor profile
version: 1
updated: <YYYY-MM-DD>

## Environment
agent: <name>
sub_agents: <yes|no>
shell: <yes|no>
cli_tools: [<detected list>]
graph_tool: <none|name>
browser_tool: <none|name>

## Defaults
report_language: <fr|en|...>
severity_scale: <default|custom, see below>
frameworks: [<list>]
workspace_root: <path>

The profile holds only questions with two legitimate answers and a skill that reads the answer. The invariants have no switch: evidence for every finding, a read-only audited repository, no exploitation detail in remediation prompts, coverage limits always reported, active testing only with the owner's recorded authorization.

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.