The Agent Skills Directory

[SAFE]: The skill implements a robust governance framework for plan inspection.
[SAFE]: It utilizes standard file system operations (std:fs.read, std:fs.write) to access and update plan documents (PLAN.md, implementation_plan.md) within the project scope.
[SAFE]: The protocol includes mandatory human-in-the-loop checkpoints, requiring explicit user confirmation of the plan artifact and individual phase findings before any amendments are applied or execution continues.
[SAFE]: No remote downloads, external dependencies, or sensitive file access patterns were found. All proposed behaviors (summarization, gap assessment, quality review) are standard AI agent tasks performed on local context.

plan-deep-dive