apply-architecture-playbook
Pass
Audited by Gen Agent Trust Hub on Aug 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed as a read-only advisory workflow. It explicitly instructs the agent not to modify code, documents, or tracker states, and avoids performing any destructive or state-changing operations like committing or pushing code.
- [EXTERNAL_DOWNLOADS]: The skill references an external repository for technical documentation regarding a PostgreSQL extension (
pg_durable). This reference targets a well-known service (GitHub) and a trusted organization (Microsoft). These references are used solely for revalidating dated facts about the extension's maturity and capabilities, which is a standard safety practice in architectural design. - [PROMPT_INJECTION]: No malicious prompt injection patterns were detected. The instructions include a clear 'Authority and composition' section that prioritizes legal requirements, safety, and explicit user direction over the playbook's own preferences, which serves as a safeguard against behavioral overrides.
- [COMMAND_EXECUTION]: There are no shell commands, subprocess calls, or dynamic execution patterns present in the skill instructions or associated markdown files.
- [DATA_EXFILTRATION]: The skill does not access sensitive files (e.g., credentials, SSH keys) or perform network operations beyond referencing public documentation for informational purposes.
Audit Metadata