One URL. Every agent remembers.
Add Axonio to any MCP client and it gets 44 memory tools: remember, search, branch, query the knowledge graph, reflect and sync.
Replace YOUR_API_KEY with a key from Dashboard → API keys.
claude mcp add --transport http axonio https://api.axonio.xyz/api/v1/mcp --header "Authorization: Bearer $AXONIO_API_KEY"Tested with
- Claude
- Cursor
- Windsurf
- GitHub Copilot
- VS Code
- OpenAI Agents
- Gemini
- LangChain
- CrewAI
- Vercel AI SDK
- Ollama
- Hugging Face
- ROS 2
- NVIDIA Isaac
Pick your client. Copy one block.
Set up your AI agent
Connect an AI agent to Axonio over the Model Context Protocol.
In Claude, find the setting for adding a custom MCP server.
Paste this configuration:
{
"mcpServers": {
"clawdb": {
"url": "https://api.axonio.xyz/api/v1/mcp",
"headers": { "Authorization": "Bearer YOUR_API_KEY" }
}
}
}Create a key from your workspace, drop it in as the Bearer token above, and your agent has memory.
44 tools, generated from the server itself.
This list is built from the MCP server’s own tool registrations, so it is always what your agent will actually see.
Memory
- clawdb_remember
- Store information in the agent's persistent memory database.
- clawdb_remember_bulk
- Store many memories in one call for fast context ingestion.
- clawdb_search
- Search the agent's memory database by meaning or keywords.
- clawdb_recall
- Retrieve specific memories by ID.
- clawdb_update_memory
- Update an existing memory entry by ID.
- clawdb_delete_memory
- Delete a memory entry by ID.
- clawdb_list_memories
- List recent memory entries.
Branching
- clawdb_branch_list
- List all memory branches.
- clawdb_branch_get
- Get a branch by ID.
- clawdb_branch_trunk
- Get the trunk (main) branch.
- clawdb_branch_fork
- Fork the agent's memory state for experimentation.
- clawdb_branch_merge
- Merge an experimental branch back into main memory.
- clawdb_branch_diff
- Diff two branches.
- clawdb_branch_discard
- Discard a branch permanently.
- clawdb_branch_archive
- Archive a branch.
Knowledge graph
- clawdb_remember_code
- Store a code snippet as memory and extract its import/class/function structure into the knowledge graph (tree-sitter based: TypeScript, JavaScript, Python).
- clawdb_remember_git_history
- Store pasted `git log`/`git show`/`git diff` output as memory and extract commit/file/author triples.
- clawdb_remember_document
- Store a Markdown document (README, ADR, runbook, API guide) as memory and extract its structure (sections/links/env-vars) and semantics (deprecates/requires/supersedes, via LLM) into the knowledge graph.
- clawdb_remember_dependencies
- Store a dependency manifest (package-lock.json contents, or a requirements.txt) as memory and extract package/version/dependency triples.
- clawdb_remember_build_log
- Store raw build/test/runtime output as memory and extract failing test / stack trace / error triples, so future failures can be matched against past ones.
- clawdb_remember_transcript
- Store a chat/meeting/email transcript as memory. An LLM extracts concrete decisions, approvals, complaints, and constraints into the knowledge graph (no formal grammar to parse, unlike the other clawdb_remember_* tools).
- clawdb_query_graph
- Query the cross-source knowledge graph (code, git history, docs, dependencies, build logs, transcripts) by entity name, relation type, and/or source type.
Reflection
- clawdb_reflect
- Trigger a reflection job.
- clawdb_reflect_list_jobs
- List reflection jobs.
- clawdb_reflect_get_job
- Get a reflection job by ID.
- clawdb_reflect_facts
- Get extracted facts.
- clawdb_reflect_preferences
- Get extracted preferences.
- clawdb_reflect_contradictions
- Get contradictions.
- clawdb_reflect_resolve_contradiction
- Resolve a contradiction.
Transaction
- clawdb_tx_begin
- Begin a memory transaction.
- clawdb_tx_remember
- Add a memory to a transaction.
- clawdb_tx_remember_typed
- Add a typed memory to a transaction.
- clawdb_tx_commit
- Commit a transaction.
- clawdb_tx_rollback
- Roll back a transaction.
Sync
- clawdb_sync
- Full bidirectional sync.
- clawdb_sync_push
- Push local changes.
- clawdb_sync_pull
- Pull remote changes.
- clawdb_sync_reconcile
- Reconcile divergent sync state.
- clawdb_sync_status
- Get sync status.
Object storage
- clawdb_store_object
- Store a small binary/text object (up to the inline limit bytes) as base64. For larger files, request a presigned upload URL out of band via the workspace REST API.
- clawdb_get_object
- Retrieve a stored object as base64.
- clawdb_list_objects
- List objects stored in this workspace.
- clawdb_delete_object
- Delete a stored object by key.
Status
- clawdb_status
- Check ClawDB connection and component health.
Call the server from this page.
> Initializing Axonio MCP session... > Connected to workspace instance [us-east-1] > Ready. Try clawdb_remember, clawdb_search, or clawdb_branch_fork.
Try it: remember a couple of facts, then recall a related query. Results are ranked from what you just stored.
Memory plumbing you write once.
One memory, every client
Per-workspace tenancy
Streamable HTTP or stdio
What is MCP?
The Model Context Protocol is an open standard for connecting AI applications to tools and data. Axonio is an MCP server: any MCP client can read and write agent memory through it without custom integration code.
Which clients work?
Any spec-compliant MCP client: Claude (Desktop and Code), Cursor, Windsurf, VS Code, ChatGPT connectors, OpenAI Agents SDK, LangChain and custom agents. Robots and background workers can use the same endpoint.
Can several clients share one memory?
Yes. Point them at the same workspace with the same or different API keys. Keys can be scoped, so a read-only client can’t write.
Can I run it locally over stdio?
Yes. npx -y @clawdb/mcp-adapter runs the MCP server over stdio on your own machine, against a local ClawDB engine (CLAWDB_URL, default http://localhost:50050). For Axonio Cloud, point the client at the hosted URL instead.