The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on executing local Git commands to manage the task lifecycle. It performs branch creation (git checkout -b), status monitoring (git status), history retrieval (git log), merging (git merge), and tagging (git tag).- [PROMPT_INJECTION]: The skill is exposed to indirect prompt injection as it processes untrusted data from task files and Git commit messages. * Ingestion points: Reads markdown files located in .wiki/tasks/(scope)/ and fetches commit logs using git log --oneline. * Boundary markers: No explicit delimiters or guardrail instructions are used to distinguish between user data and agent instructions within the task files. * Capability inventory: The skill has the authority to create and switch branches, merge code into the develop branch, create annotated tags, and modify files within the repository. * Sanitization: There is no evidence of sanitizing task names or descriptions before they are used to generate Git branch names or commit messages, which could lead to command manipulation if the agent handles shell characters improperly.

make-tasks