[Laurianti]

Documentation only: ChromaDotNet.Embeddings in the docs of the client.

docs/embeddings.md starts with ChromaDotNet.Embeddings, which computes the vectors of the default embedding function of Chroma, the ones of the Python client, before the example with OpenAI, which stays:

• a short example: ChromaDefaultEmbeddingGenerator.CreateAsync(), a ChromaCollectionDefinition that declares the function default, WithEmbeddingGenerator, AddAsync with documents only and QueryAsync by text;
• one sentence on the limits: a preview, 64-bit processes only, not on Intel Macs, with the link to its repository;
• the SearchByText sample of ChromaDotNet.Local and its Codespace.

The README links it next to the line of ChromaDotNet.Local.

The example was built and run as written, against ChromaDotNet.Client 2.18.0 and ChromaDotNet.Embeddings 1.0.0-preview.3 from nuget.org, on Chroma 1.5.9: "something sour" finds the lemon, and the collection stores {"type":"known","name":"default"} as its embedding function.

No code changes, no new version.

[Copilot]

🟢 Approval recommended

The documentation matches the client APIs and linked package documentation, with no unresolved issues.

No open issues 🎉

What changed in this PR

Documents ChromaDotNet.Embeddings as a way to generate Python-compatible default embeddings in .NET.

Changes:

• Adds a README link to the embeddings package and guide.
• Adds a usage example, platform limitations, and sample/Codespace links while retaining the OpenAI example.

File	Description
README.md	Links to ChromaDotNet.Embeddings and its guide.
docs/​embeddings.md	Introduces default embeddings with setup, usage, limitations, and sample links.

_{Review effort: Balanced}