The Agent Skills Directory

[SAFE]: The skill is entirely descriptive and outlines a multi-step process for conformance testing and analysis improvement. It does not include any executable code or configuration that would lead to unauthorized actions.- [SAFE]: No indicators of prompt injection, data exfiltration, or obfuscation were found. The skill does not access sensitive files or perform network communications.- [SAFE]: The instruction to run a review command refers to a workflow step within the Fallow analysis environment and does not involve risky execution patterns or privilege escalation.

conformance-loop