improving-prompts
Pass
Audited by Gen Agent Trust Hub on Oct 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process existing LLM prompt text or files provided by the user.
- Ingestion points: The skill accepts input via the
argument-hintas either direct text or a file path. - Boundary markers: The process lacks explicit delimiters or instructions to ignore potential commands embedded within the untrusted prompt text being improved.
- Capability inventory: The skill has the capability to read files from the local filesystem and edit files in place when requested by the user.
- Sanitization: There is no evidence of input validation or sanitization to ensure the input content does not contain instructions targeting the agent performing the improvement.
Audit Metadata