plan
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingest untrusted user data in the form of implementation requests or specifications to drive repository analysis and planning.
- Ingestion points: The main workflow in
SKILL.mdis triggered by a user-provided 'request or specification'. - Boundary markers: The instructions do not specify the use of delimiters or protective framing to isolate untrusted input from the agent's core instructions.
- Capability inventory: The skill involves reading the repository state (
references/_shared/orchestration/reality-check.md) and writing implementation plans (references/_shared/artifacts/plan.md), creating a read-write capability chain. - Sanitization: There is no mention of sanitizing or validating the input specification for malicious payloads or hidden instructions.
Audit Metadata