triage

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill contains a meta-instruction to override the agent's current configuration by explicitly demanding a model change ("First set the /model to Haiku").
  • [COMMAND_EXECUTION]: The skill executes destructive file system operations, specifically the automated deletion of files within the todos/ directory when a user selects the "next" option during triage.
  • [COMMAND_EXECUTION]: The workflow incorporates system-level commands and references custom CLI tools (e.g., ls and /resolve_todo_parallel) to manage and interact with local project files.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 03:33 PM