create-agentsmd
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill requires the agent to ingest and analyze untrusted project data (such as source code, build configurations, and CI/CD files) to generate documentation. This creates a surface where malicious instructions embedded in the analyzed files could influence the agent's behavior.
- Ingestion points: Analysis of project structure,
package.jsonscripts, Makefiles, build files, and CI/CD configuration files. - Boundary markers: The prompt lacks specific instructions or delimiters to tell the agent to ignore potentially malicious instructions found within the analyzed files.
- Capability inventory: The agent is expected to perform file system operations (writing
AGENTS.md) and is encouraged to "test the instructions by ensuring all commands work," which implies shell command execution capabilities. - Sanitization: No sanitization or validation of the ingested file content is specified in the prompt.
Audit Metadata