Agentic HDF5

A collection of tools and skills that enable AI agents to work at a high level with HDF5 data and files.

Included Skills:

• HDF-optimization Skill: Skill teaching agents best-practice for HDF5 chunking, chunk cache usage, filter/compression usage, and read/write practices

Included Tools:

• Tools for creation and parsing of Semantic Metadata (SMD) attributes to track data provenance, significance, and other high-level scientific attributes. See docs/semantic-metadata.md for details
• Tools for creation and querying of Vectorized Semantic Metadata (VSMD), enabling natural-language/keyword based queries over data within SMD attributes. See docs/vectorized_smd_design.md for details