agentleFS
VulnetixVulnetix/pix-ai-coding-assistant/skills/repo-impact/SKILL.md

repo-impact

Skill9 starsChanged 7 days agoApache-2.0View on GitHub

Answer whether a specific advisory reaches anything in THIS repository — which manifest declares the package, at what version, whether the version is in range, and whether the vulnerable code is reachable. Use when someone asks "does CVE-… affect us", when triaging an advisory against a real codebase, or before spending effort on a fix.

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:

    If a scan was already recorded, .vulnetix/memory.yaml has it, and vulnetix agent hook puts a summary in front of the agent at session start.

    line 63

≈ “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

Repository impact

Use when

  • "Does CVE-2021-44228 affect us?" — the only question that matters when an advisory is in the news.
  • Triaging an advisory that names a package this repository might pull in transitively.
  • Deciding whether a fix is worth the effort before starting it.

Don't use for

  • Looking the advisory up. That is vulnetix_vuln over MCP, or vulnetix vdb vuln <id>, and it needs no skill. This skill exists for the half that needs the working tree.
  • Applying the fix — fix.
  • Confirming a fix landed — verify-fix.

Conventions

Follows skills/_lib/contract.md for surface selection, output style and memory writes.

What "affects us" actually means

Four questions, in order, and each one can end the enquiry:

  1. Is the package here at all? Directly, or pulled in transitively.
  2. Is the installed version in the affected range? Being on the package is not being vulnerable.
  3. Is there a fixed version, and how far away is it? A patch release and a major rewrite are different answers.
  4. Is the vulnerable code reachable? Present in the tree is not the same as called.

Report which of these you could answer and which you could not. An unreachable finding and an unanswered reachability question look identical in a summary and mean opposite things.

Step 1: Ask the tree, not the internet

vulnetix scan --sca -o json

This resolves the actual dependency graph, including transitives, and anchors each package to the manifest line that declares it. Grep is not a substitute: it misses transitives entirely, which is where most advisories land.

If a scan was already recorded, .vulnetix/memory.yaml has it, and vulnetix agent hook puts a summary in front of the agent at session start. Reuse that when it is current, and say it is a snapshot.

Step 2: Get the advisory's own facts

vulnetix vdb vuln "$ARGUMENTS" -o json

Or vulnetix_vuln with detail: summary over MCP, which is about 10 KB rather than megabytes.

Take the affected ranges, the fixed versions, the affected routines, and whether it is known-exploited.

Step 3: Intersect

Match the advisory's affected ranges against the versions actually resolved in step 1. Be exact about ranges — >=2.0.0 <2.17.1 does not include 2.17.1, and an off-by-one here is the difference between an emergency and a non-event.

Step 4: Reachability, when the CLI can answer it

vulnetix scan --reachability -o json

The advisory names affected routines; reachability says whether this repository calls them. When it cannot answer — no analyser for the language, a dynamic call graph — say so rather than reporting UNKNOWN as if it were UNREACHABLE.

Step 5: Answer the question that was asked

<CVE-…>: <AFFECTED | NOT AFFECTED | CANNOT DETERMINE>

  Package    <name>@<installed>
  Declared   <manifest>:<line>   (<direct | transitive via X>)
  Affected   <range>  ->  fixed in <version>
  Reachable  <yes: path | no | not analysed for <language>>
  Exploited  <known exploited | no evidence>

Lead with the verdict. If the answer is NOT AFFECTED, say why in one line — wrong version, wrong platform, not present — because "no" without a reason gets asked again next week.

Record the verdict in .vulnetix/memory.yaml so the next session, and dashboard, can see the decision rather than repeat the work.

Next: fix <CVE-…> when the answer is AFFECTED.

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.