write-implementation-plan
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process potentially untrusted data such as project specifications, instructions, and code, which could contain malicious embedded instructions.
- Ingestion points: The skill reads specifications, project instructions, and existing code or tests as part of its 'Preflight' process.
- Boundary markers: While it requires an 'approved' behavioral source, it lacks explicit delimiters or instructions to ignore nested prompt injections within the source material.
- Capability inventory: The skill generates executable implementation plans that include validation commands and test obligations, creating a path for malicious instructions to influence system operations.
- Sanitization: The instructions do not specify any sanitization or escaping of the content retrieved from external specifications before including it in the generated plan.
Audit Metadata