skills/rheinmir/setup/council/Gen Agent Trust Hub

council

Pass

Audited by Gen Agent Trust Hub on Jul 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local script harness/scripts/council.py to perform deterministic protocol math, including anonymization via SHA256, mean-rank aggregation, and report rendering.
  • [COMMAND_EXECUTION]: It utilizes the orca orchestration CLI to manage model task creation, dispatching, and result collection across the multi-stage workflow.
  • [PROMPT_INJECTION]: The skill processes untrusted content from multiple LLM agents (Stage-1 answers and Stage-2 reviews). This creates an indirect prompt injection surface; however, the skill documentation explicitly specifies that all input strings are sanitized using html.escape before being included in the mandatory Stage-4 HTML reports to prevent injection attacks.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 18, 2026, 03:07 PM
Security Audit — agent-trust-hub — council