The Agent Skills Directory

[SAFE]: The skill integrates exclusively with LangWatch's official tools and infrastructure. It uses the langwatch CLI for all operations, which is the standard interface for the vendor. All external documentation and data endpoints (e.g., langwatch.ai) are official vendor resources. The instructions include security best practices, such as warning the agent to never print, copy, or send API keys, and advising that keys should be read directly from the project's environment files. It also correctly distinguishes between shared organizational projects and personal workspaces to prevent data misalignment. The error reporting feature (npx langwatch report) includes explicit user-approval flags and local scrubbing of secrets before submission.

online-evaluations