implement
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute verification commands derived directly from the task documents (
docs/task/*.md). While these are intended for testing the implementation, they represent a mechanism for arbitrary command execution within the agent's environment. - [INDIRECT_PROMPT_INJECTION]: The skill relies on external markdown files as the primary source of instructions for code changes and verification steps, creating a surface for indirect prompt injection.
- Ingestion points: The agent reads
docs/task/*.md,TASKS.md, andAGENTS.mdto define its actions (Workflow steps 1-4). - Boundary markers: The skill includes a check requiring the task to be reviewed and approved before implementation begins (Workflow step 5).
- Capability inventory: The agent is granted permission to modify arbitrary files listed in the task and execute shell commands for verification (Workflow steps 8-9).
- Sanitization: No specific filtering or sanitization of the task document content is performed; the skill relies on manual approval and the agent's safety assessment.
Audit Metadata