skills/knoopx/pi/jj-hunk/Gen Agent Trust Hub

jj-hunk

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill documents the use of the jj-hunk command-line utility for managing Jujutsu version control hunks.
  • It provides examples for running commands such as jj-hunk list, jj-hunk split, jj-hunk commit, and jj-hunk squash.
  • These commands are intended for local version control management and do not exhibit suspicious behavior.
  • [DATA_EXFILTRATION]: No patterns associated with credential theft or network exfiltration were detected.
  • The skill uses standard temporary file paths like /tmp/spec.yaml for configuration and does not access sensitive directories such as .ssh or .aws.
  • [REMOTE_CODE_EXECUTION]: No remote code execution patterns, such as piping network requests to a shell, were found.
  • All commands described are local to the user's environment and the Jujutsu repository.
  • [PROMPT_INJECTION]: The skill does not contain any instructions that attempt to override AI safety guidelines or extract system prompts.
  • [EXTERNAL_DOWNLOADS]: No instructions were found that download or install external packages or scripts at runtime.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 09:19 AM
Security Audit — agent-trust-hub — jj-hunk