loop-me
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes content from
NOTES.mdand markdown files within theworkflows/directory to sharpen terminology and generate workflow specifications. Untrusted data within these files could potentially influence the agent's behavior during the interactive session. However, the agent's capabilities are limited to local file system modifications within a specific workspace. - Ingestion points:
NOTES.md,workflows/*.md(SKILL.md) - Boundary markers: None explicitly defined in the instructions to separate processed data from instructions.
- Capability inventory: File creation, editing, and deletion within the
workflows/directory andNOTES.md(SKILL.md). - Sanitization: None detected; the skill relies on the model to interpret and record terminology directly from raw notes.
Audit Metadata