interrogate
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute
git diff main...HEADto identify code changes for review. This is a standard and appropriate use of development tools for this skill's purpose.\n- [PROMPT_INJECTION]: The skill processes untrusted code changes which are interpolated into sub-agent prompts, creating a surface for indirect prompt injection.\n - Ingestion points: Code diffs or file contents retrieved in
SKILL.mdStep 1.\n - Boundary markers: The sub-agent prompt uses a markdown header to separate code, but lacks strict delimiters or instructions to ignore embedded prompts within the code block.\n
- Capability inventory: Parent agent can
spawn_agentandreadfiles; sub-agents are restricted to read-only tasks.\n - Sanitization: No validation or sanitization is performed on the code content before processing.
Audit Metadata