agentleFS
mtarcuremtarcure/claude-vibe-squad/.agents/skills/known-advisory-backport-check/SKILL.md

known-advisory-backport-check

Skill220 starsChanged 20 days agoMITView on GitHub

Use when a target forks or pins a dependency (a forked cosmos/evm, an old cosmos-sdk/cometbft, a vendored library) — enumerate pinned versions, pull the upstream advisory set, and find known vulnerabilities the target failed to backport. Cheapest high-value pass on any fork.

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:

    Enumerate pinned versions: Read go.mod/Cargo.toml/package.json including replace/=> directives (the replace wins — audit the ACTUAL …

    line 12

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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

known-advisory-backport-check

For any target that forks or pins a dependency (a forked cosmos/evm, an old cosmos-sdk/cometbft, a vendored library): known upstream vulnerabilities the target failed to patch are reachable, low-effort findings. Run this early — it's the cheapest high-value pass and it's what the 52-competitor crowd's osv-scan finds first.

Steps

  1. Enumerate pinned versions. Read go.mod/Cargo.toml/package.json including replace/=> directives (the replace wins — audit the ACTUAL resolved version, e.g. cosmos/evm => a bridge target/evm v1.0.0-rc2...).
  2. Pull the advisory set. osv-scanner --lockfile <lock> for the dep graph; gh api repos/<upstream>/security-advisories for published GHSAs; the upstream SECURITY.md / release notes / advisory index (Hacken, Cosmos ASA, etc.).
  3. For each Critical/High advisory: get affected + first_patched_version (gh api /advisories/<GHSA>), compare to the pinned version. Behind the fix = candidate.
  4. Reachability + config. Is the vulnerable module/function actually WIRED and reachable at no privilege in THIS target's config? (e.g. is x/group registered + in EndBlock? is the ICS20 precompile activated? is the affected code path called?) A vuln in an unused code path doesn't pay.
  5. Fork-diff for missing backports. If it's a fork of a release-candidate (-rcN) or a frozen branch, the upstream may have shipped a security fix AFTER the fork point. Diff the fork vs upstream for the fix's presence — a code-pattern fix (e.g. a snapshot/journal wrap, a signer check) can be silently dropped in a merge even when the version looks recent. Confirm the guard is present at the pinned commit.

The scope/dedup reality (be honest — protect the rep)

  • Known public CVEs are usually dedup-dead and often ruled "known / out-of-scope / third-party" — osv-scan is the crowd's first move. Do NOT submit a stock reachable-CVE as-is expecting payment; it's a near-certain rejection + rep penalty.
  • What DOES convert: (a) a fix that is genuinely MISSING/INCOMPLETE in the fork at the pinned commit (novel to this deployment), or (b) a Push-/target-SPECIFIC elevation where custom code makes the generic bug reachable/impactful in a way the advisory didn't cover. Novelty + a runnable PoC is the bar.
  • Either way: report the reachable-known ones to the project as free remediation (goodwill), and only SUBMIT the novel/incomplete-fix ones. Feed the decision to impact-validator (G3 dedup, G4 scope). Related: [[chain-construct]].

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.