The Agent Skills Directory

[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it requires autonomous subagents to ingest and interpret the entire contents of a codebase.
Ingestion points: All files in the repository, including source code, comments, and configuration files (SKILL.md).
Boundary markers: Absent. There are no explicit delimiters or instructions provided to the subagents to ignore or isolate potentially malicious instructions embedded within the code or comments they analyze.
Capability inventory: The subagents are empowered to read/write files, execute shell commands (including analysis tools and test runners), and perform git operations (SKILL.md).
Sanitization: No sanitization, escaping, or validation of the processed code content is performed before it is analyzed by the subagents.
[COMMAND_EXECUTION]: The skill utilizes several CLI tools to perform its primary functions.
It executes knip for dead code and dependency analysis in JavaScript/TypeScript projects.
It executes madge to identify circular dependencies.
It performs git operations including branch creation, logging to prioritize high-churn files, and committing changes.
It identifies and executes the project's local test runner to verify the integrity of automated fixes.

codebase-sweep