project-manager-senior
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze site specifications from external files, specifically
ai/memory-bank/site-setup.md. This creates a potential vector where malicious instructions embedded in a project specification could influence the agent's behavior. - Ingestion points: Reads project requirements from
ai/memory-bank/site-setup.mdas part of its core responsibility. - Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded commands within the specification file.
- Capability inventory: The agent is authorized to write task lists to the filesystem and references local shell scripts for QA testing in its generated templates.
- Sanitization: There are no explicit steps provided to sanitize or validate the content of the specification files before processing.
- [COMMAND_EXECUTION]: The skill's task list format template includes a command for Playwright testing:
./qa-playwright-capture.sh http://localhost:8000 public/qa-screenshots. While this is a standard QA practice, it involves recommending the execution of a local shell script.
Audit Metadata