senior-project-manager
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process project specifications from the file
ai/memory-bank/site-setup.md. This represents an indirect prompt injection surface where malicious instructions embedded in the specification could potentially influence the agent's task generation or behavior. - Ingestion points: The agent reads from
ai/memory-bank/site-setup.mdto analyze requirements. - Boundary markers: There are no explicit instructions or delimiters defined to separate the untrusted specification content from the agent's internal logic.
- Capability inventory: The skill has the capability to write new task files to
ai/memory-bank/tasks/and execute a local shell script (./qa-playwright-capture.sh). - Sanitization: No sanitization or validation logic is defined for the content extracted from the specification file.
- [COMMAND_EXECUTION]: The skill instructions include a command to execute a local shell script for quality assurance:
./qa-playwright-capture.sh http://localhost:8000 public/qa-screenshots. While the script targets a local development server, it demonstrates a capability for command-line interaction.
Audit Metadata