The Agent Skills Directory

[SAFE]: No security issues detected. The skill consists entirely of instructional markdown files providing guidance on using the Jujutsu (jj) version control tool and the GitHub CLI (gh).
[COMMAND_EXECUTION]: The skill instructs the agent to execute jj and gh commands. It includes necessary safeguards for an AI environment, such as mandating the -m flag to prevent commands from hanging on interactive text editors and forbidding interactive commands like jj split.
[PROMPT_INJECTION]: The skill contains instructions for an "immutable_heads bypass." This is a legitimate feature of the Jujutsu tool used to rewrite protected history. The skill includes a explicit instruction: "Always ask the user before disabling immutable protection," ensuring the user retains control over high-risk version control operations.

using-jj