cavecrew
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues detected. The skill provides instructions for managing subagent delegation and defines structured output formats to optimize context window usage.
- [INDIRECT_PROMPT_INJECTION]: The skill establishes an indirect prompt injection surface by delegating tasks to subagents whose output is injected verbatim into the main context.
- Ingestion points: Subagent tool results are injected into the main thread (SKILL.md, README.md).
- Boundary markers: The skill defines structured 'output contracts' (e.g., `:
- path:line — symbol — short note`) to help distinguish data from instructions.
- Capability inventory: Subagents are authorized to locate code, perform surgical edits (1-2 files), and review diffs (SKILL.md).
- Sanitization: No explicit sanitization is described, though the 'Auto-clarity' section prioritizes safety warnings over compression.
Audit Metadata