skills/danielmiessler/lifeos/Council/Gen Agent Trust Hub

Council

Pass

Audited by Gen Agent Trust Hub on Aug 2, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the curl utility to send a POST request to http://localhost:31337/notify. This is used as a local notification mechanism to inform the system when a workflow is initiated.
  • [COMMAND_EXECUTION]: The skill records its activity by appending JSON entries to a local log file located at ~/.claude/LIFEOS/MEMORY/SKILLS/execution.jsonl using standard shell redirection.
  • [DATA_EXPOSURE]: The skill scans for and loads configuration files (PREFERENCES.md) from the user's local directory at ~/.claude/LIFEOS/USER/CUSTOMIZATIONS/SKILLS/Council/ to apply custom behavior overrides.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it interpolates user-provided topics directly into the system prompts of multiple sub-agents. However, the use of structured round instructions and specialized personas serves as a functional boundary for these interactions.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 2, 2026, 08:46 AM
Security Audit — agent-trust-hub — Council