one-shot
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user prompts to guide the implementation and verification phases without implementing safety boundaries or sanitization.
- Ingestion points: The initial user task description is ingested in Step 1 (Assess ceremony) and used to generate implementation plans.
- Boundary markers: The skill does not instruct the agent to use delimiters or specific ignore-rules to isolate user instructions from the agent's execution logic.
- Capability inventory: The skill allows writing implementation logs to the filesystem (
thoughts/directory), delegating tasks to sub-agents, and executing shell commands (Step 4). - Sanitization: No sanitization or validation is performed on the user prompt before it influences code generation and command execution.
- [COMMAND_EXECUTION]: The workflow requires the agent to execute shell commands defined during the planning phase to verify the implementation. This capability allows for arbitrary command execution based on the task prompt provided by the user.
Audit Metadata