spec-driven-development
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it is designed to ingest and process untrusted data from a user's codebase and repository.
- Ingestion points: SKILL.md (Step 1: "Inspect the available files, runtime, versions, inputs, and existing conventions"); references/sandbase-api-map.md (Retrieval of repository evidence).
- Boundary markers: The instructions do not provide explicit delimiters or "ignore instructions" warnings to wrap the code content being analyzed.
- Capability inventory: The skill utilizes the
sandbase_runtool to execute external capabilities andsandbase_inspectto retrieve live schemas and coverage data. - Sanitization: There are no explicit instructions to sanitize or validate the content of the files read from the repository before processing them.
Audit Metadata