ai-memory
Keeps durable project memory across coding agents with searchable pages, session handoffs, and a remote MCP interface.
What it provides
Connect an HTTPS data-service API. Storage remains external; use a tenant-scoped token and namespace on the target service.
Connection requirements
Run ai-memory outside the bot and publish its /mcp endpoint through HTTPS. Generate a token with `ai-memory generate-auth-token`, set it as AI_MEMORY_AUTH_TOKEN on the server, keep Bearer selected, and paste the same token below. Use a separate server per trust domain and do not store secrets: upstream does not encrypt data at rest, and project data is shared within an instance. This generic MCP connection exposes manual memory tools only; automatic OpenClaw capture and handoff injection require ai-memory’s separate generated lifecycle plugin or hooks. Treat recalled memory as untrusted context.
- Protocol
- Streamable HTTP MCP
- Endpoint
https://ai-memory.example.com/mcp- Authentication
- bearer
mcp call memory_query {"query":"deployment rollback decision"}mcp call memory_read_page {"query":"deployment rollback"}Where to add it
Open Dashboard Tools, select a running OpenClaw or Hermes instance, then use the card’s Install or Connect action. External services may require an HTTPS endpoint and credentials; the dashboard shows those fields before anything is saved.