architect
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes bash commands that interpolate path variables (e.g., {vocabulary.notes}) read from local configuration files (ops/derivation-manifest.md). This presents a potential command injection surface if the configuration values contain shell metacharacters.
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes untrusted data from the filesystem to generate recommendations.
- Ingestion points: Files in {vocabulary.notes}/, ops/observations/, and ops/sessions/ are read and processed (SKILL.md).
- Boundary markers: No delimiters or safety warnings are specified for the ingested content to prevent instructions from being obeyed.
- Capability inventory: The skill utilizes Bash, Write, Edit, and specialized search tools like mcp__qmd__deep_search (SKILL.md).
- Sanitization: No sanitization or filtering of external content is mentioned before the data is processed or presented to the user.
Audit Metadata