The Agent Skills Directory

[SAFE]: The skill consists solely of natural language instructions (markdown) and a configuration file. It does not include any scripts, executable binaries, or remote code patterns.
[SAFE]: The instructions explicitly guide the agent to respect privacy boundaries, avoid public disclosure of sensitive source materials, and require explicit user approval before making any external changes or mutations (e.g., GitHub, email, public posts).
[SAFE]: The skill implements a 'Value Checkpoint' which requires the agent to pause for approval before modifying project state if the user's intent is ambiguous, serving as a safeguard against unintended behavior.

source-intake