The Agent Skills Directory

[SAFE]: The skill instructions define a workflow for agentic evaluation using SandBase tools (sandbase_discover, sandbase_inspect, sandbase_run, etc.). These tools are part of the vendor's infrastructure (sandbaseai) and are used for their intended purpose.
[SAFE]: The skill includes explicit security and budget controls, such as requiring user confirmation for paid calls, minimizing data sent to external APIs, and prioritizing local work over remote calls.
[SAFE]: No obfuscation, prompt injection, unauthorized data access, or malicious persistence mechanisms were detected in the instructions or reference files.

agentic-eval