enhance-agent-prompts
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted agent prompt files to perform analysis and automated modifications.
- Ingestion points: The skill reads and parses .md files from platform-specific agent directories (e.g., ~/.claude/agents/, ~/.config/opencode/agents/).
- Boundary markers: No specific delimiters or instructions are provided to the agent to treat the contents of the analyzed files as untrusted data or to ignore instructions embedded within them.
- Capability inventory: The skill instructions enable the agent to find, read, parse, and write modifications (auto-fixes) to files on the local filesystem.
- Sanitization: No evidence of sanitization or validation of the analyzed file content is present before the agent processes and acts upon the data.
Audit Metadata