agents-consilium
Pass
Audited by Gen Agent Trust Hub on May 9, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: Orchestrates tasks by executing established CLI tools including codex, claude, opencode, and gemini. These interactions are confined to plan-only or read-only modes to prevent unauthorized system changes.
- [DATA_EXFILTRATION]: Facilitates the transmission of user-specified code and prompts to reputable AI service providers (Google, Anthropic, OpenAI) as part of its core functionality.
- [PROMPT_INJECTION]: Mitigates risks by utilizing CDATA sections and escaping sequences to ensure user-provided content does not interfere with the orchestration logic or sub-agent instructions.
- [SAFE]: No malicious patterns such as obfuscation, persistence mechanisms, or unauthorized privilege escalation were detected. The skill follows security best practices by implementing least-privilege access for its sub-agents.
Audit Metadata