The Agent Skills Directory

[COMMAND_EXECUTION]: The skill constructs and executes shell commands using the dub and git CLI tools. It dynamically generates command strings by interpolating variables like <branch-name> and <commit-message> which are derived from the repository's current state and user interaction.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it ingests and analyzes data that could be controlled by external actors (e.g., commit messages or code changes from other contributors).
Ingestion points: The skill reads untrusted data via git diff --cached and git log --oneline during Phase 1 (Analyze Changes).
Boundary markers: No explicit delimiters or instructions are provided to the agent to treat the output of Git commands as data rather than instructions.
Capability inventory: The skill has access to execute file system operations and repository modifications through the git and dub command-line interfaces.
Sanitization: There is no evidence of sanitization or filtering of the content retrieved from the Git repository before it is processed by the AI to suggest branch names and commit summaries.

dub-flow