jj-todo-workflow

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION] (SAFE): The file documents the use of Jujutsu (jj) commands for workspace management. While it includes rm -rf for cleaning up workspace directories, the usage is contextually appropriate for the described workflow and targeted at specific user-created paths.
  • [PROMPT_INJECTION] (SAFE): The instructions provide templates for guiding agents to work in specific directories. These do not contain patterns designed to bypass AI safety guardrails or override core system instructions.
  • [DATA_EXPOSURE] (SAFE): No hardcoded credentials, API keys, or access patterns for sensitive system files (such as SSH keys or cloud provider configs) were found.
  • [INDIRECT_PROMPT_INJECTION] (LOW): The workflow involves agents reading task descriptions from revision metadata. This is a standard data ingestion surface for task-oriented agents and does not present an immediate security risk in the context provided.
  • [EXTERNAL_DOWNLOADS] (SAFE): No network requests, remote script downloads, or external package installations are initiated by this skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:13 PM