api-design-rest
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses a local Python utility,
scripts/validate_api_contract.py, to ensure that API manifests comply with defined governance rules such as ID formats and state transitions. This script is executed locally by the agent as part of the quality gate process. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, including product requirements and resource model candidates, which creates a potential surface for indirect prompt injection attacks.
- Ingestion points: Requirements and manifest files are ingested during the modeling and validation phases in
SKILL.mdandreferences/api-governance-contract.md. - Boundary markers: The instructions lack explicit boundary delimiters or warnings to treat ingested design data as potentially untrusted.
- Capability inventory: The skill includes the ability to execute the local Python validation script as defined in the workflow.
- Sanitization: While the Python script validates JSON structure and governance logic, it does not sanitize descriptive or natural language fields for prompt injection patterns.
Audit Metadata