The Agent Skills Directory

[SAFE]: The skill consists entirely of instructional markdown and does not include any executable code, network-based exfiltration, or hardcoded credentials. It promotes security best practices such as using isolated workspaces and requiring human confirmation for critical actions.
[NO_CODE]: No scripts, binaries, or external code dependencies are included or downloaded by this skill.
[COMMAND_EXECUTION]: The skill defines a process for the agent to execute implementation tasks and verifications. This behavior is mediated by a mandatory 'critical review' step in the workflow where the agent is instructed to identify and raise concerns before proceeding with execution.

executing-plans