The Agent Skills Directory

[SAFE]: The skill incorporates comprehensive data redaction guidelines to prevent the exposure of secrets, absolute system paths, and PII in project documentation. It explicitly lists types of sensitive data to avoid and provides placeholder strategies for necessary references.
[SAFE]: Operational capabilities are limited to reading project files and executing scoped development tools (such as git, npm, and pytest) to verify documentation claims. These commands are constrained to their primary project roles as defined in the platform's tool configuration.
[SAFE]: No obfuscation, remote code execution vectors, or unauthorized network activities were identified. The skill includes mandatory self-review protocols and assumption-checking to ensure that documentation remains accurate and safe before being written to disk.

write-for-ai