The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses git, gh (GitHub CLI), and pandoc to manage repositories and process documents. These operations are limited to the author's own infrastructure (browseros-ai on GitHub) and are protected by safety guards that verify environment state before execution.- [SAFE]: No signs of obfuscation, credential theft, or unauthorized network communication were detected. The skill implements strong isolation by performing all write operations in a deterministic temporary directory (/tmp/internal-docs-<slug>) rather than the user's primary checkout.- [SAFE]: The skill includes defensive instructions to prevent accidental changes to critical files like .gitmodules and requires explicit validation of the working directory before running git commands.- [SAFE]: The use of disable-model-invocation: true in the frontmatter provides an additional layer of security by preventing the agent from autonomously calling other skills while processing potentially untrusted local documentation content.

write-internal-docs