skill-council

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill mandates the execution of a shell script (orchestrate.sh) using the bash interpreter to run the council workflow.
  • Evidence: "${CLAUDE_PLUGIN_ROOT:-$HOME/.claude-octopus/plugin}/scripts/orchestrate.sh" council <user arguments> in SKILL.md.
  • [INDIRECT_PROMPT_INJECTION]: The skill acts as a pass-through for user-supplied input into a shell command, creating a potential command injection surface.
  • Ingestion points: User-provided <user arguments> passed to the orchestrate.sh script.
  • Boundary markers: None identified in the execution command to delimit user input from shell arguments.
  • Capability inventory: Shell command execution via bash.
  • Sanitization: No specific sanitization or validation of the <user arguments> is described in the skill instructions prior to execution.
  • [DATA_EXPOSURE]: The skill accesses the user's environment to resolve the path for its plugin scripts, specifically targeting the home directory.
  • Evidence: Reference to $HOME/.claude-octopus/plugin.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 02:45 AM
Security Audit — agent-trust-hub — skill-council