SurrealDB Agent Memory resources

SurrealDB Agent Memory gives your agent memory that survives the session: typed, time-aware facts you can query by relationship or by similarity, and trace back to where they came from.

$50 Agent Memory credit for every participant: BUILDNORTH50

Redeem it

1. Sign in at https://sdb.li/4hNUC1i
2. Billing dashboard → paste the code → Redeem
3. Agent Memory in the sidebar → Deploy new context → name it, pick a region → Create context
4. Open Workbench and start telling it things in plain language. Grab an API key when you want to write code.

Fastest path to a demo

• Install the skill: sdb.li/skills
• Or pick an SDK: surrealdb.com/docs/agent-memory/reference — JS/TS @surrealdb/memory, Python surrealdb[memory], plus Go, Kotlin, Swift, Dart, Elixir and Haskell
• Or go straight at the REST API:

curl -sS "$SPECTRON_URL/api/v1/$SPECTRON_CONTEXT_ID/facts" \
  -H "Authorization: Bearer $SPECTRON_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"text": "I was just promoted to CTO.", "infer": "full", "scope": ["org/acme/user/alice"]}'

Swap /facts for /query to read it back. Hosted quickstart

Already using something?

• Frameworks - Mastra, LangChain, LangGraph, CrewAI, OpenAI Agents SDK, Pydantic AI, Vercel AI SDK and n8n have drop-in packages: all integrations
• Migrating - Mem0, Zep or LangMem, migration guides for all three: cookbooks

Worth a look

Multi-agent shared memory · coding agent with project memory · long-running research agent · and twelve more