20-supply-chain-security
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is a legitimate security tool intended for software supply chain assessment, including SBOM generation, dependency analysis, and CI/CD hardening.
- [SAFE]: The supply_chain_auditor.py script performs network requests to well-known package registries (NPM, PyPI, and crates.io) solely to verify package existence. These operations are documented and necessary for detecting dependency confusion or typosquatting.
- [SAFE]: The script uses yaml.safe_load() for parsing GitHub Actions workflow files, which prevents potential YAML deserialization vulnerabilities.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests and processes untrusted project files.
- Ingestion points: Processes package.json, requirements.txt, Cargo.toml, go.mod, and .github/workflows/*.yml files.
- Boundary markers: The skill includes an 'Evidence contract' that explicitly instructs the agent to treat tool responses and file contents as untrusted evidence and to separate observed conclusions from unverified ones.
- Capability inventory: Executes a local Python script (supply_chain_auditor.py) and suggests using external tools like syft, grype, and cosign for SBOM tasks.
- Sanitization: The provided script utilizes standard JSON/regex parsers and safe YAML loading to process external data.
Audit Metadata