dl-discover
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied requirement details and trade-offs, which creates a potential surface for indirect prompt injection.\n
- Ingestion points: User inputs during the multi-round discussion phase described in
SKILL.mdsteps 2 and 3.\n - Boundary markers: The instructions do not define specific delimiters (e.g., XML tags or triple quotes) or provide warnings for the agent to ignore instructions embedded within the user's requirement descriptions.\n
- Capability inventory: The skill has the capability to read local filesystem contents (specifically
../double-loop/references/dependencies.mdinSKILL.md) and write summary files to the filesystem (step 4).\n - Sanitization: There is no mention of sanitizing, escaping, or validating the requirement summaries before they are written to persistent storage.
Audit Metadata