agentleFS
nassaj2011nassaj2011/claude-skills-for-chatgpt/skills/dependency-supply-chain-auditor/SKILL.md

dependency-supply-chain-auditor

Skill0 starsChanged 26 days agoMITView on GitHub

Trace dependency inventory, lockfile provenance, vulnerabilities, licenses, install-time scripts and CI action permissions with evidence-backed remediation.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If authorized tools and databases are available, run read-only SBOM, vulnerability and license scanners.

    line 15
  2. Distinctive:

    No unauthorized installs, silent upgrades, fabricated CVE matches, covert telemetry or unsupported license/legal conclusions.

    line 27

≈ “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

Dependency and supply-chain audit — provenance before scores

Appropriate use

Apply to package upgrades, dependency threats, license approval, SBOM, registry trust, container base images, CI marketplace actions and lockfile integrity. Not a general code style check.

Steps

  1. Record exact source SHA and actual ecosystem manifests/lockfiles, package manager, runtime platform and environment. Do not guess dependency versions or introduce packages without permission.
  2. Build direct/transitive inventory: package | pinned version/source | dependency path | runtime/dev | integrity/provenance | license/SPDX | repo evidence.
  3. Identify registry substitution/typosquatting, dynamic dependency ranges, install/postinstall scripts, git-sourced tarballs, compromised CI actions, unverified binary downloads and overbroad workflow tokens. Use pinned trusted references where possible.
  4. If authorized tools and databases are available, run read-only SBOM, vulnerability and license scanners; capture scanner/advisory database timestamps, severities, fixed ranges, false positives and coverage. Never fabricate live CVEs or current advisory conclusions.
  5. Assess each potential finding for reachability, configuration, exposure, vulnerable call path and exploit prerequisites. CVSS alone is not deploy impact. Maintain separate VERIFIED / SUSPECTED / UNKNOWN.
  6. Propose minimal upgrades with compatibility test matrix, safe lockfile regeneration plan, rollback and provenance checks; do not update installed dependencies automatically or alter public registry contents.
  7. Gate critical issues on ownership, exact manifest change, actual rerun evidence and approval or documented exception with an expiry. A clean static lint check cannot certify the entire supply chain.

Required output

Inventory + SBOM plan | suspicious source and script checks | evidence-linked risk table | fixed-range plan | license obligations | CI permissions recommendations | residual blind spots.

Runnable reference

examples/supply-chain/ implements a toy lockfile policy linter against synthetic dependency metadata. It flags floating ranges, non-registry sources and lifecycle script concerns but is NOT an npm/pip parser or live vulnerability feed. In ChatGPT Web without repository access, ask for redacted manifest excerpts and write commands to execute later, marked NOT RUN.

Limitations

No unauthorized installs, silent upgrades, fabricated CVE matches, covert telemetry or unsupported license/legal conclusions.

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.