agents-orchestrator
Pass
Audited by Gen Agent Trust Hub on Aug 10, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill acts as a process manager for development tasks. No malicious patterns, obfuscation, or credential theft attempts were detected. The agent's instructions focus on systematic quality gates and progress tracking.
- [COMMAND_EXECUTION]: The skill utilizes basic Unix utilities such as 'ls', 'cat', and 'grep' to manage project files and verify task completion status. These operations are limited to project-specific directories (e.g., project-specs/, project-tasks/) and do not target sensitive system locations or configuration files.
- [DATA_EXFILTRATION]: No network activity, such as 'curl' or 'wget' to external domains, was observed. The skill does not access or transmit sensitive environment variables or credentials.
- [PROMPT_INJECTION]: The skill's instructions are purely operational and do not include attempts to bypass safety filters, override system prompts, or extract internal instructions. It includes internal guardrails such as 'quote EXACT requirements' to ensure fidelity to project specifications.
Audit Metadata