blueprint-execute
Pass
Audited by Gen Agent Trust Hub on May 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill performs state detection through read-only shell commands on local project files located in the docs/ directory.
- [COMMAND_EXECUTION]: Shell operations are focused on reading file counts, checking versions, and parsing JSON data. For file updates, the skill follows best practices by using 'jq --arg' for variable injection, which effectively prevents command injection vulnerabilities.
- [SAFE]: The skill ingests untrusted data from project documentation such as PRDs and ADRs (Ingestion points: docs/prds, docs/adrs, manifest.json; Boundary markers: absent; Capability inventory: Bash, SlashCommand; Sanitization: jq). While this constitutes an ingestion surface, the risk is mitigated as the operations are local and lack network access or sensitive file exposure.
- [SAFE]: No obfuscation, persistence mechanisms, or credential harvesting patterns were found in the skill body or reference documentation.
Audit Metadata