The Agent Skills Directory

[SAFE]: The skill operates entirely on the local filesystem, scanning JSON artifacts and status files within the project's .plan directory. This behavior is restricted to the workspace and is appropriate for the skill's diagnostic purpose.
[COMMAND_EXECUTION]: The tool executes internal project utilities such as 'manage-lessons list' to verify that referenced lesson IDs exist in the live inventory. These are trusted calls to project infrastructure.
[DATA_EXFILTRATION]: There is no evidence of data exfiltration. The skill does not perform network operations to untrusted domains, and all diagnostic findings are recorded in the local project store.

plan-doctor