analyze-project
Pass
Audited by Gen Agent Trust Hub on Apr 4, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill identifies entrypoints and code patterns in a repository, which presents a surface for indirect prompt injection.
- Ingestion points: The
scripts/analyze_project.pyscript reads directory structures and file contents from the target repository path provided via the--repoargument. - Boundary markers: The generated reports (
SUMMARY.md,RISKS.md) do not utilize delimiters or specific instructions to isolate untrusted repository metadata from the agent context. - Capability inventory: The skill possesses file system read and write capabilities for local analysis but lacks network access or subprocess execution of the analyzed content.
- Sanitization: Filenames and structural metadata are included in the final analysis output without escaping or validation, allowing crafted project structures to influence the agent.
Audit Metadata