context-architecture
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, secrets, or dangerous commands were identified in the skill. The instructions are purely methodological and promote security best practices such as least privilege through explicit boundaries and automated verification.
- [PROMPT_INJECTION]: The skill requires the agent to audit untrusted repository documentation (e.g., README, AGENTS.md) and suggests verifying that commands within those files 'still run'. This identifies an indirect prompt injection surface inherent to the tool's primary purpose.
- Ingestion points: Repository documentation files including README, AGENTS.md, and CLAUDE.md as described in Phase 1 and Phase 3.
- Boundary markers: Absent; the skill does not specify the use of delimiters to wrap external content during the audit.
- Capability inventory: The skill requires the agent to perform file system reads and execute shell commands to verify documentation accuracy.
- Sanitization: Absent; the methodology relies on the agent to interpret and verify documentation content without specific sanitization instructions.
Audit Metadata