The Agent Skills Directory

[COMMAND_EXECUTION]: The skill performs various file system and repository operations using the Bash tool.
Evidence: Executes git worktree, git branch, git push, grep, mkdir, and gh pr create commands throughout SKILL.md.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to the way it handles external repository data and user input.
Ingestion points: Data is ingested into the agent context via git worktree list results, .gitignore file content, and user-provided branch descriptions in SKILL.md sections 2, 3, and 5.
Boundary markers: Absent. There are no clear delimiters or instructions telling the agent to treat data from the repository as untrusted.
Capability inventory: The skill has access to the Bash, Read, and Write tools, allowing for command execution and file modification.
Sanitization: Absent. The instructions do not define any validation or escaping for the branch names or command outputs before they are processed or executed.

worktree-manager