gang-skeptic
Pass
Audited by Gen Agent Trust Hub on May 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
hive teamcommand to fetch environment data to identify its specific instance and team name. - [PROMPT_INJECTION]: The skill is designed to process inputs from other agents, which constitutes an indirect prompt injection surface. Evidence: 1. Ingestion points: Receives feature plans and validation verdicts from peer agents. 2. Boundary markers: None defined. 3. Capability inventory: Internal messaging via
hive send. 4. Sanitization: Relies on the agent's skepticism instructions to evaluate and challenge the integrity of the data it receives.
Audit Metadata