The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill generates a Python script by substituting user-provided variables into a markdown template and then executes this script using the uv tool to validate its functionality.
[EXTERNAL_DOWNLOADS]: The generated synchronization script is designed to download documentation archives (zip or tar.gz) from remote URLs provided by the user or inferred from the environment.
[COMMAND_EXECUTION]: The skill executes various system commands and developer tools, including ruff, mypy, pyright, and uv, to format, lint, type-check, and run the generated code.
[INDIRECT_PROMPT_INJECTION]: The skill processes external documentation which is subsequently indexed in a SKILL.md file, creating a vector for instructions embedded in remote data to influence agent behavior.
Ingestion points: Documentation archives are downloaded from the user-specified DOC_SOURCE_URL and extracted locally.
Boundary markers: The processed documentation titles and descriptions are placed within a designated ## Documentation Index section in the target SKILL.md file.
Capability inventory: The skill and its generated script have the capability to write to the filesystem, execute shell commands, and perform network requests.
Sanitization: The workflow includes "markdown grooming" to remove Hugo shortcodes and transform links for AI consumption, but it does not perform explicit sanitization of natural language instructions within the documentation.

add-doc-updater