ultraqa
Pass
Audited by Gen Agent Trust Hub on Apr 11, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes project-defined shell commands for testing, building, and linting as part of its core QA workflow. It also performs file system management by deleting its own local state files upon completion.
- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it incorporates command outputs and build logs into subagent prompts without explicit sanitization or boundary markers.
- Ingestion points: Command output and test results are passed to the architect subagent in SKILL.md.
- Boundary markers: No delimiters are defined to separate ingested log content from instructions.
- Capability inventory: The skill is capable of file modification and shell command execution.
- Sanitization: No sanitization of logs or output is performed before analysis.
Audit Metadata