The Agent Skills Directory

[SAFE]: The skill implements a research-oriented workflow that manages experiment routing and metric provenance without external network calls or sensitive data access.
[SAFE]: The instructions reference local project documentation (e.g., ../satisfy-your-agent/references/experiments.md) to establish context, which is a standard method for modular agent skills to share knowledge.
[SAFE]: Explicit ethical constraints are provided, such as instructions to not translate behavioral preferences into consciousness claims and to store no raw model responses, demonstrating a focus on data privacy and alignment.

sya-research