cargo-vet-audit

name	cargo-vet-audit
description	Orchestrates cargo-vet supply chain audits for Rust crates. Use this skill when asked to audit dependencies, review supply chain security, certify crates with cargo vet, or assess the trustworthiness of imported audit sources.

Cargo-Vet Audit Skill

You are orchestrating a cargo vet supply chain audit. Follow this process end-to-end.

Step 1: Discover Unvetted Crates

Run cargo vet and parse the output. Omit --locked only if imports.lock needs to be refreshed for imported third-party audits; this does not refer to updating Cargo.lock. Categorize each unvetted crate as either:

• Delta audit — a version-to-version diff (e.g., 1.0.0 → 1.1.0)
• Full audit — a complete source inspection of a single version

Note the recommended commands from cargo vet's output (e.g., cargo vet diff, cargo vet inspect).

Step 2: Plan the Audit

Present the user with a table of all unvetted crates:

Crate	Type	Audit Size	Notes
...	delta/full	files/lines	...

Ask the user to confirm before proceeding.

Step 3: Delegate to the Cargo-Vet Auditor Agent

For each crate, delegate the actual code review to the cargo-vet-auditor agent. Launch multiple agents in parallel when there are many crates to audit.

Provide each agent with:

• The crate name and version(s)
• Whether this is a delta or full audit
• The exact command to run (cargo vet diff CRATE FROM TO or cargo vet inspect CRATE VERSION)
• The working directory

Step 4: Compile Results

Collect agent results into a confidence score table:

Crate	Type	Unsafe	Build/Proc Macro	Powerful Imports	Advertised Behavior	Confidence	Verdict

Confidence scoring rubric:

• 95-100: No unsafe, no build script, no powerful imports, trivial/well-known crate
• 90-94: Minimal unsafe (sound, reviewed), simple build script, well-understood crate
• 80-89: Non-trivial unsafe (sound but complex), build script with FS access, larger crate
• 70-79: Complex unsafe requiring careful review, proc macros with code generation
• 60-69: Concerns noted but mitigated, unusual patterns
• Below 60: Red flags found — do NOT certify, escalate to user

Step 5: Certify

Exemptions Are a Last Resort

Do not add [[exemptions]] entries without explicit user confirmation. Each exemption bypasses the audit process entirely and must be justified. Valid reasons include:

• The crate is only needed for safe-to-run (test/dev tooling) and a full audit is disproportionate
• An upstream import source was removed and the crate needs temporary coverage while a first-party audit is scheduled
• The user explicitly requests an exemption after reviewing the trade-offs

Always prefer auditing (full or delta) over exempting. When an exemption is unavoidable, present it to the user for manual approval before adding it.

Every exemption must include a --notes explaining why the exemption exists and under what conditions it can be removed:

cargo vet add-exemption CRATE VERSION --criteria CRITERIA \
  --notes "Reason for exemption; plan for resolution"

For each crate that passes (confidence ≥ 70), run:

cargo vet certify CRATE FROM TO --accept-all --criteria safe-to-deploy \
  --who "NAME <EMAIL>" --notes "AUDIT_NOTES"

For full version audits (no delta), omit the FROM version:

cargo vet certify CRATE VERSION --accept-all --criteria safe-to-deploy \
  --who "NAME <EMAIL>" --notes "AUDIT_NOTES"

Use the git user's name and email for --who.

AI Attribution in Audit Notes

Following the Linux kernel's AI attribution guidelines, every audit note MUST include an Assisted-by tag to transparently disclose that the audit was performed with AI assistance. Use the format:

Assisted-by: AGENT_NAME:MODEL_ID cargo-vet

Where:

• AGENT_NAME is copilot-cli (or the specific agent framework)
• MODEL_ID is the model that performed the review (e.g., claude-sonnet-4.5, claude-opus-4.6). Determine this from the session's model configuration.
• cargo-vet is the specialized analysis tool used

For example, a complete --notes value would be:

"No unsafe, no build script, no I/O. Assisted-by: copilot-cli:claude-opus-4.6 cargo-vet"

The human user remains responsible for reviewing all AI-generated audit assessments and certifications. The --who field must always identify the human reviewer, never the AI agent.

Step 6: Verify and Clean Up

Before final verification, detect and remove identical duplicate [[audits.*]] entries that may have been appended by retried cargo vet certify commands.

Duplicate-check workflow:

1. Scan supply-chain/audits.toml for byte-for-byte identical audit blocks
2. If duplicates exist, keep one copy (usually the first) and remove the rest
3. Re-run cargo vet after deduplication to ensure state is still valid

To detect duplicates, scan supply-chain/audits.toml for repeated blocks with identical crate name, who, criteria, version/delta, and notes fields. Remove any duplicates before proceeding.

Then run the normal cleanup sequence:

4. Run cargo vet again to confirm everything passes
5. Run cargo vet prune to remove stale exemptions
6. Run cargo vet one final time to confirm clean state

Reviewing Import Sources

When asked to review imported audit sources (in supply-chain/config.toml), evaluate each on:

Factor	Weight	How to Assess
Organization reputation	High	Known security-conscious org? (Mozilla, Google, Bytecode Alliance, etc.)
Activity / freshness	High	Last commit date, commit frequency
Community size	Medium	Stars, forks, contributors
Audit coverage	Medium	Number of unique crates audited
Domain relevance	Medium	Does their audit focus overlap with our dependency graph?
Dedicated audit repo	Low	Dedicated repo vs. audits inside a product repo

Present results as a confidence score table with reasoning.