agentleFS
poliakarmaipoliakarmai/gsc/skills/gsc-triage/SKILL.md

gsc-triage

Skill17 starsChanged yesterdayApache-2.0View on GitHub

Use when triaging GSC findings — blocking set, TP vs noise, PoC verification.

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:

    Coverage first: If detectors_not_run is non-empty or scan_status is partial, the finding set is incomplete — say so before discussing …

    line 37

≈ “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

GSC Triage

When to use

You have GSC findings (from the Action, the CLI, or MCP) and must decide what blocks a merge, what is noise, and which findings a PoC actually confirms.

Commands

MCP tools (server gsc, shipped inside the engine image):

list_findings     # recent findings from the local DB, no scan
get_finding       # full detail for one finding_key
get_poc           # the PoC attached to a finding (DB or deterministic template)
verify_finding    # re-run one PoC in the sandbox
verify_findings   # same for a batch of keys
doctor            # self-check: sandbox, state volume, allowed roots, cache, lock
list_detectors    # which detectors the engine has

CLI context helpers:

docker run --rm -v "$PWD:/repo:ro" -w /repo ghcr.io/poliakarmai/gsc-scanner:v1 sca --repo .
docker run --rm ghcr.io/poliakarmai/gsc-scanner:v1 epss --cve CVE-XXXX-YYYY

How to read the result

Triage in this order:

  1. Coverage first. If detectors_not_run is non-empty or scan_status is partial, the finding set is incomplete — say so before discussing findings.
  2. Blocking set. The Action decides it via fail_on_critical and fail_on_score (threshold 1–100). Findings that make the gate exit non-zero are the blocking subset; everything else is advisory.
  3. Verified = confirmed. verify_finding returning an exploit status means the PoC reproduced — treat the finding as confirmed and fix or explicitly accept it.
  4. unavailable is not "not reproducible". It means the sandbox could not start (missing user-namespace flags, no PoC attached, wrong volume). Run doctor before drawing any conclusion.
  5. No PoC ≠ false positive. Secrets in particular have no deterministic template; a live credential stays a true positive without any PoC.
  6. inconclusive stays open. Report it as "static match, unverified" rather than dismissing it.

Pitfalls

  • list_findings reads the DB, not your working tree: confirm it is the same repo run you are triaging (get_finding re-scans only when the key is absent).
  • Batch with verify_findings instead of looping verify_finding.
  • verify_fix with status fixed means "the detector no longer fires after a re-scan" (GSC-003), not "the exploit was executed and failed". Do not overclaim.
  • DESTRUCTIVE and non-deterministic PoCs are refused by design; a refusal is a policy verdict, not a statement about the vulnerability.

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.