product-interface-spec
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run a local Python validation script,
scripts/validate_delivery.py, to verify the integrity and structure of generated documentation. Analysis of the script shows it only uses safe standard libraries (argparse, json, pathlib) for structural checks and contains no network or dangerous code execution capabilities. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided external product specifications and legacy documents, which represents an attack surface for indirect prompt injection where malicious instructions could be embedded in source data.
- Ingestion points: Reads existing project specifications, source files, and design documents as specified in
SKILL.mdandreferences/delivery-contract.md. - Boundary markers: Enforces the use of fixed Markdown templates and JSON schema validation, but lacks explicit delimiter-based sanitization for natural language content.
- Capability inventory: Ability to write Markdown/JSON files and execute the local Python-based structural validator.
- Sanitization: Performs structural and reference integrity validation via script, but does not perform semantic sanitization of the natural language content from external sources.
Audit Metadata