The Agent Skills Directory

[SAFE]: The skill provides meta-instructions designed to enhance the agent's accuracy and reliability during technical tasks. It mandates practices such as re-reading code before editing, running verification tests, and avoiding API hallucinations.
[SAFE]: No indicators of data exfiltration, credential theft, or unauthorized network access were found. The skill does not attempt to execute remote code or install external dependencies.
[SAFE]: The instructions do not contain prompt injection techniques aimed at bypassing safety filters. Instead, they reinforce strict adherence to logical reasoning and verification protocols.

zero-defect