canon
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use a project-specific CLI tool named 'doctrine' to perform searches and display specifications, architecture decision records (ADRs), policies, and standards. This is standard functionality for retrieving local documentation.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest data from external files such as 'CLAUDE.md' and 'AGENTS.md', as well as the output of the 'doctrine' tool. While these serve as ingestion points for potentially untrusted data that lacks explicit boundary markers or sanitization in the skill instructions, the ingestion is necessary for the skill's primary purpose of informing the agent of project governance. The agent's capabilities include memory retrieval and consultation tools which are used to process implementation details correctly.
Audit Metadata