optimize-agents-md
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest, analyze, and rewrite user-controlled instruction files (
AGENTS.md), which creates a surface for indirect prompt injection. - Ingestion points: The skill reads
AGENTS.mdandAGENTS.override.mdfrom the workspace root or specified paths. - Boundary markers: The skill lacks explicit boundary markers or instruction-ignoring delimiters when processing the content of these target files.
- Capability inventory: The skill is authorized to perform file-write operations to modify the
AGENTS.mdfile based on its analysis. - Sanitization: While the instructions tell the agent to verify facts against the repository, there is no programmatic sanitization or escaping of the ingested text to prevent it from containing adversarial instructions that might influence the agent's behavior during the optimization process.
Audit Metadata