ask-matt
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [NO_CODE]: The skill is composed exclusively of markdown documentation and metadata configuration. It does not contain any executable logic, scripts (such as Python, Node.js, or Shell), or binary files that could perform operations on the host environment.\n- [INDIRECT_PROMPT_INJECTION]: The skill describes workflows designed to ingest and process untrusted data from external sources, which is a standard surface for indirect prompt injection. This analysis identifies the potential attack surface inherent in the described software development lifecycle.\n
- Ingestion points: Untrusted data enters the agent's context through bug reports and feature requests in the
/triageflow, as well as through primary source investigation in the/researchflow.\n - Boundary markers: While the skill mentions context hygiene and management tools like
/compactand/handoff, no explicit delimiters or specific instructions for isolating untrusted data are provided in these high-level routing files.\n - Capability inventory: The system documented by this router includes capabilities for code writing (
/implement), behavior testing (/tdd), and an interactive setup for credentials and secrets (/wizard).\n - Sanitization: There is no evidence of input validation, escaping, or sanitization logic within the provided routing instructions.
Audit Metadata