Get started

Honcho for AI agents

Honcho provides memory and context infrastructure for stateful AI agents, tracking peers, sessions, and conclusions over time. Product and engineering teams use it as a building block inside their own AI features. 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.

Connect HonchoSee the 16 actions ↓
AI and modelsAPI keyMCP + RESThoncho.dev
AUDIT LOG · HONCHOPOLICY: acme-support
09:41:07 · claude · u_8f2read
honcho.list_peers
List Peers✓ allowed · 212ms
09:41:08 · claude · u_8f2read
honcho.get_message
Get Message✓ allowed · 164ms
09:41:09 · claude · u_8f2write
honcho.create_messages
Create Messages✓ approved · approved by user
09:41:12 · claude · u_8f2destructive
honcho.delete_session
Delete Session✕ blocked · policy: deny
EVERY HONCHO CALL, ON THE RECORD
01 · USE CASES

What agents do in Honcho.

01

Recall context for a peer

Pull a peer's stored representation to give an agent context before continuing a conversation.

02

Log a new session message

Save a new message to a session so the conversation history stays complete.

03

Search past conclusions

Search stored conclusions for a topic to reuse a prior insight instead of recomputing it.

02 · ACTIONS

16 Honcho actions, graded by risk.

Every Honcho action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.

read

12

Look things up. Allowed by default.

  • honcho.list_peers
    List Peers
  • honcho.get_message
    Get Message
  • honcho.list_messages
    List Messages
  • honcho.list_sessions
    List Sessions
  • honcho.list_workspaces
    List Workspaces
  • honcho.search_messages
    Search Messages
  • honcho.list_conclusions
    List Conclusions
  • honcho.get_or_create_peer
    Get or Create Peer
  • honcho.search_conclusions
    Search Conclusions
  • honcho.get_or_create_session
    Get or Create Session
  • honcho.get_or_create_workspace
    Get or Create Workspace
  • honcho.get_peer_representation
    Get Peer Representation

write

3

Create and change things. Allow, or ask the user first.

  • honcho.create_messages
    Create Messages
  • honcho.create_conclusions
    Create Conclusions
  • honcho.chat_with_peer
    Chat With Peer

destructive

1

Delete, cancel or archive. Ask first, or deny outright.

  • honcho.delete_session
    Delete Session
03 · HOW IT WORKS

Honcho in three steps.

  1. 01Your users connect HonchoThey add their Honcho api key on Arc0 Connect, under your brand. It goes straight into the vault.
  2. 02You set the rulesReads run, writes like “create Messages” can wait for the user, and “delete Session” can be denied outright.
  3. 03Any agent can actYour agent calls Honcho through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
POLICY.TS
await arc0.policies.set('honcho', {
  read: 'allow',
  write: 'ask',        // create_messages
  destructive: 'deny',  // delete_session
})

# Claude Code: the same connection, one URL
$ claude mcp add --transport http arc0 \
    https://mcp.arc0.ai/u/u_8f2
04 · AUTH AND DATA

How Honcho connects.

Users add their Honcho 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 Honcho 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
05 · WORKS WITH

Use Honcho from any agent.

Claude
ChatGPT
Cursor
Codex
VS Code
OpenAI Agents SDK
Claude Agent SDK
Vercel AI SDK
Mastra
LangGraph
06 · RELATED

More ai and models apps.

Perplexity AI9 actionsAffinda119 actionsAgent Mail5 actionsAI/ML API30 actionsAivoov1 actionsAll Images AI7 actionsAltText.ai10 actionsAltTextLab2 actions
ALL AI AND MODELS APPS →
07 · FAQ

Honcho and Arc0, answered.

Q01

Can I use Honcho with Claude, ChatGPT or Cursor?

Yes. Connect Honcho 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 Honcho actions you allow.

Q02

How do users connect Honcho?

Users add their Honcho 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.

Q03

Which Honcho actions can my agent take?

16 in total: 12 read, 3 write and 1 destructive, such as “create Messages”. Your policies decide which of them each agent may call.

Q04

Can I stop my agent from deleting things in Honcho?

Yes. Actions like “delete Session” 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.

Q05

Can my own backend call Honcho too?

Yes. The same Honcho connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.

Get started

Plug Honcho into your agent.

Your users connect Honcho once, under your brand. Your agent gets 16 actions behind your policies, with every call on the record.

Connect HonchoBrowse all apps →

Free to build · MCP + REST · Audit log on every plan