create-agentsmd
Pass
Audited by Gen Agent Trust Hub on May 15, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill uses instructional language to direct the agent's behavior and define its task. These instructions are focused on generating documentation and do not attempt to bypass safety guidelines.\n- [COMMAND_EXECUTION]: The skill directs the agent to "Test the instructions by ensuring all commands work as documented," which creates a vulnerability to indirect prompt injection if the analyzed repository contains malicious commands.\n
- Ingestion points: Analyzes project files such as package.json, Makefile, and CI/CD configurations as outlined in the implementation steps.\n
- Boundary markers: None. The prompt does not include instructions to limit or sanitize the execution of commands found in the repository.\n
- Capability inventory: The instruction implies a requirement for the agent to execute shell commands to verify project setup and testing workflows.\n
- Sanitization: No validation or sanitization of repository-provided commands is mentioned.\n- [EXTERNAL_DOWNLOADS]: The skill references https://agents.md/ for guidance and examples regarding the AGENTS.md format.
Audit Metadata