The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use standard Git CLI commands such as git worktree, git merge, and git push for managing isolated development environments. These operations are appropriate and expected for the skill's stated purpose.
[PROMPT_INJECTION]: The workflow involves agents reading data from external sources like CLI arguments and shared files, which creates a surface for indirect prompt injection. This is a fundamental part of the intended coordination mechanism. 1. Ingestion points: CLI --message arguments, PRD files, and response files as described in the testing protocol section of SKILL.md. 2. Boundary markers: None specified in the instructions to separate untrusted data from system instructions. 3. Capability inventory: File system read/write access and git subprocess execution across the lifecycle. 4. Sanitization: No sanitization or validation logic is defined for the content of the shared messages or files.

shared-worktree