The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill downloads and installs the official claat tool from a trusted GitHub repository associated with the Google organization.
[EXTERNAL_DOWNLOADS]: The skill fetches and executes claat-tools directly from the author's (nogikun) repository using the go run command.
[COMMAND_EXECUTION]: The skill invokes shell commands such as go run, go install, and claat to manage its build environment and perform document linting and conversion.
[DYNAMIC_EXECUTION]: The skill utilizes go run to compile and execute Go code from a remote repository at runtime. This mechanism is the primary way the skill provides its linting and building capabilities and involves code maintained by the skill author.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied Markdown files (manual.md), which constitutes an ingestion surface for untrusted data.
Ingestion points: Reading and linting the content of the manual.md file (found in SKILL.md).
Boundary markers: No explicit delimiters or boundary markers are defined to isolate the content of the processed files.
Capability inventory: The skill possesses the capability to execute shell commands (go run, claat) and write files to an output directory.
Sanitization: The skill relies on the internal parsing and validation logic of the claat-tools and claat binaries to safely handle the Markdown input.

claat-build