Langbase for AI agents
Langbase is a serverless platform for developers to build and deploy AI agents, with memory, pipes, and threads as building blocks. Connect it once through Arc0, and your agent, or Claude, ChatGPT and Cursor, can use it through one MCP endpoint, limited to what each user approved.
What agents do in Langbase.
Add a message to a thread
Append a new message to an active conversation thread run through a pipe.
Look up execution traces
Read recent execution traces for a pipe to debug an agent's behavior.
Approve before deleting memory
Require approval before deleting a memory store, since it removes indexed documents an agent relies on.
20 Langbase actions, graded by risk.
Every Langbase action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.
read
6Look things up. Allowed by default.
- langbase.get_pipeGet Pipe Details
- langbase.get_threadGet Thread Details
- langbase.list_modelsList Available Models
- langbase.list_tracesList Execution Traces
- langbase.list_thread_messagesList Thread Messages
- langbase.check_healthCheck API Health
write
11Create and change things. Allow, or ask the user first.
- langbase.update_pipeUpdate an existing pipe
- langbase.create_threadCreate Thread
- langbase.update_threadUpdate Thread Metadata
- langbase.update_thread_messageUpdate Thread Message
- langbase.pipe_listList all pipes
- langbase.memory_listList Memories
- langbase.pipe_createCreate a new pipe
- langbase.document_listList Documents in Memory
- langbase.memory_createCreate Memory
- langbase.chunker_split_contentSplit Content into Chunks
- langbase.append_thread_messagesAppend Thread Messages
destructive
3Delete, cancel or archive. Ask first, or deny outright.
- langbase.delete_threadDelete Thread
- langbase.delete_thread_messageDelete Thread Message
- langbase.memory_deleteDelete Memory
Langbase in three steps.
- 01Your users connect LangbaseThey add their Langbase api key on Arc0 Connect, under your brand. It goes straight into the vault.
- 02You set the rulesReads run, writes like “create Thread” can wait for the user, and “delete Thread” can be denied outright.
- 03Any agent can actYour agent calls Langbase through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
await arc0.policies.set('langbase', { read: 'allow', write: 'ask', // create_thread destructive: 'deny', // delete_thread }) # Claude Code: the same connection, one URL $ claude mcp add --transport http arc0 \ https://mcp.arc0.ai/u/u_8f2
How Langbase connects.
Users add their Langbase api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
The same Langbase connection serves your agent over MCP and your own backend over REST and the proxy, so a user connects once. How Arc0 handles credentials →
- AUTH
- API key
- CREDENTIALS
- Per-tenant encrypted vault
- MODEL SEES
- Results only, never credentials
- AUDIT LOG
- Every call, on every plan
Use Langbase from any agent.
Langbase and Arc0, answered.
Can I use Langbase with Claude, ChatGPT or Cursor?
Yes. Connect Langbase to Arc0 once, then add your Arc0 MCP URL to Claude, ChatGPT, Cursor, Claude Code or any other remote-MCP client. Each assistant only gets the Langbase actions you allow.
How do users connect Langbase?
Users add their Langbase api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
Which Langbase actions can my agent take?
20 in total: 6 read, 11 write and 3 destructive, such as “create Thread”. Your policies decide which of them each agent may call.
Can I stop my agent from deleting things in Langbase?
Yes. Actions like “delete Thread” are graded destructive. Set destructive actions to deny, or to ask so the user approves each one, and blocked calls still show up on the audit log.
Can my own backend call Langbase too?
Yes. The same Langbase connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.
Plug Langbase into your agent.
Your users connect Langbase once, under your brand. Your agent gets 20 actions behind your policies, with every call on the record.
Free to build · MCP + REST · Audit log on every plan