The Agent Skills Directory

[COMMAND_EXECUTION]: The skill manages development workflows by executing git commands for worktree creation and cleanup, as well as running test suites to evaluate implementation variants.- [EXTERNAL_DOWNLOADS]: Implementation logic is retrieved from an external hosted LLM service. While this is a core architectural feature, it involves the ingestion of code generated outside the local environment.- [PROMPT_INJECTION]: The skill ingests code from an external LLM, creating a surface for indirect prompt injection. Ingestion points: implementation code from hosted LLM services via tool calls. Boundary markers: no delimiters or warnings are used to signal the agent to ignore embedded instructions in generated code. Capability inventory: the skill can write files, manage git repositories, and execute tests. Sanitization: no validation or sanitization of the generated code is performed before execution.

any-percent