The Agent Skills Directory

[SAFE]: The skill implements a structured multi-step onboarding process (role, plugins, connectors, trial) using internal platform tools. No external code or unauthorized network activity is present.
[SAFE]: The instructions explicitly require the agent to call system tools (like list-plugins and list-connectors) and wait for their results before communicating with the user, which prevents state hallucinations and ensures accuracy.
[SAFE]: No obfuscation, prompt injection attempts, or persistence mechanisms were detected in the skill instructions or metadata.

setup-cowork