agents-md
Pass
Audited by Gen Agent Trust Hub on Aug 9, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill employs high-authority, imperative language (e.g., "Master Prompt", "Principal Architect", "Overwrite Authority", "Must execute") to direct agent behavior and suppress user confirmation steps ("Do not ask... immediately generate"). These instructions are designed to ensure automated execution but utilize patterns common in prompt injection to override default agent deliberation.- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow where the agent ingests untrusted data from the project environment to generate governance instructions, creating an indirect prompt injection surface.
- Ingestion points: Local project files including source code,
README.md,docs/, and configuration files (e.g.,package.json). - Boundary markers: The skill does not define specific boundary markers or delimiters for the data being analyzed to prevent embedded instructions from influencing the agent's output.
- Capability inventory: The skill utilizes file system write capabilities to create or modify
AGENTS.mdandCLAUDE.mdfiles. - Sanitization: No specific sanitization or validation logic is defined for the content extracted from project files before it is used to formulate Golden Rules.
Audit Metadata