ajeetraina777/sbx-surrealdb-kits repository overview

⁠SurrealDB kit for Docker Sandboxes

A standalone Docker Sandboxes⁠ kit (kind: mixin) that adds an embedded, multi-model SurrealDB⁠

• documents, graph edges, and native vector search - plus the surrealdb Python SDK to any sandbox agent. SurrealDB runs in-process (embedded storage engines ship with the SDK), so there is no server and no external database.

Vector search needs an embedder. This image ships in three flavors, one per tag; SurrealDB is the same in all of them and only the embedder changes.

Source and full docs: https://github.com/ajeetraina/sbx-kits-surrealdb⁠

⁠Image tags

DMR is the default because it needs no cloud keys. It matters most for Claude agents: Anthropic does not offer an embeddings model (https://docs.anthropic.com/en/docs/build-with-claude/embeddings⁠), so a Claude user has no cloud embedder. DMR fills that gap locally. OpenAI and Gemini users can reuse a key they already have.

⁠Quick start

Local default (DMR). Enable Docker Model Runner and pull the two models on the host:

docker model pull ai/gemma3
docker model pull ai/mxbai-embed-large
sbx run --kit docker.io/ajeetraina777/sbx-surrealdb-kits:latest claude

OpenAI. Store the key once with sbx (never on the command line), then run:

echo "$OPENAI_API_KEY" | sbx secret set -g openai
sbx run --kit docker.io/ajeetraina777/sbx-surrealdb-kits:openai claude

Gemini:

echo "$GOOGLE_API_KEY" | sbx secret set -g google
sbx run --kit docker.io/ajeetraina777/sbx-surrealdb-kits:gemini claude

The cloud tags hold no key. The sbx proxy injects it from the stored secret, so the key never enters the sandbox. sbx run has no -e flag by design.

⁠How it works

Each kit installs the surrealdb SDK (with embedded engines), writes /home/agent/.surrealdb/config.json with the connection and the right embedder (model + vector dimensions: DMR 1024, OpenAI 1536, Gemini 768), and adds the matching embedder domain to the sandbox allow list. The store persists on disk at ~/.surrealdb/data. Runnable demos ship under ~/runbooks/ (travel.py for vector memory, graph.py for multi-model graph queries).

Per-provider setup notes, validation details, and the raw spec.yaml for each kit live on GitHub: https://github.com/ajeetraina/sbx-kits-surrealdb/tree/main/providers⁠