inspect-code
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external source code modules, which provides a pathway for malicious instructions embedded in code comments or data to potentially influence the agent's analysis or actions.
- Ingestion points: The skill ingests target paths (files, directories, or subsystems) in 'Phase 1: Scope Triage & Macro Grounding' to perform structural audits.
- Boundary markers: The execution protocol lacks explicit instructions for the agent to treat audited file content as untrusted data or to use specific delimiters to isolate it from the agent's own instructions.
- Capability inventory: The skill utilizes file and symbol discovery tools and possesses the capability to launch multiple 'research' subagents for broader repository analysis.
- Sanitization: There is no mention of sanitizing, escaping, or filtering the ingested source code content before it is processed by the model or subagents.
Audit Metadata