Get started

Honeycomb for AI agents

Honeycomb is an observability platform for high-cardinality event data, letting engineers query traces and spans to debug production systems. Developers wire it into their own apps and internal tools instead of building the underlying capability from scratch. 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 HoneycombSee the 33 actions ↓
Developer toolsOAuth 2.0 (MCP)MCP + RESThoneycomb.io
AUDIT LOG · HONEYCOMBPOLICY: acme-support
09:41:07 · claude · u_8f2read
honeycomb.list_spans
List spans✓ allowed · 212ms
09:41:08 · claude · u_8f2read
honeycomb.get_trace
Get trace✓ allowed · 164ms
09:41:09 · claude · u_8f2write
honeycomb.create_slo
Create slo✓ approved · approved by user
EVERY HONEYCOMB CALL, ON THE RECORD
01 · USE CASES

What agents do in Honeycomb.

01

Run a query on a dataset

Run a query against a dataset to find the slowest requests in the last hour.

02

Pull a trace's spans

Fetch the spans for a specific trace to diagnose where latency is coming from.

03

Create a trigger with approval

Create a new alert trigger only after the on-call engineer confirms the threshold.

02 · ACTIONS

33 Honeycomb actions, graded by risk.

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

read

16

Look things up. Allowed by default.

  • honeycomb.get_trace
    Get trace
  • honeycomb.list_spans
    List spans
  • honeycomb.get_dataset
    Get dataset
  • honeycomb.get_signals
    Get signals
  • honeycomb.list_boards
    List boards
  • honeycomb.find_columns
    Find columns
  • honeycomb.find_queries
    Find queries
  • honeycomb.get_triggers
    Get triggers
  • honeycomb.get_environment
    Get environment
  • honeycomb.list_recipients
    List recipients
  • honeycomb.get_span_details
    Get span details
  • honeycomb.get_query_results
    Get query results
  • honeycomb.get_aiconversation
    Get aiconversation
  • honeycomb.get_dataset_columns
    Get dataset columns
  • honeycomb.list_aiconversations
    List aiconversations
  • honeycomb.get_workspace_context
    Get workspace context

write

17

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

  • honeycomb.create_slo
    Create slo
  • honeycomb.update_slo
    Update slo
  • honeycomb.create_board
    Create board
  • honeycomb.update_board
    Update board
  • honeycomb.create_marker
    Create marker
  • honeycomb.create_trigger
    Create trigger
  • honeycomb.update_trigger
    Update trigger
  • honeycomb.create_recipient
    Create recipient
  • honeycomb.semconv
    Semconv
  • honeycomb.feedback
    Feedback
  • honeycomb.run_query
    Run query
  • honeycomb.run_bubbleup
    Run bubbleup
  • honeycomb.refinery_docs
    Refinery docs
  • honeycomb.migration_guide
    Migration guide
  • honeycomb.canvas_agent_invoke
    Canvas agent invoke
  • honeycomb.record_onboarding_state
    Record onboarding state
+ 1 MORE

destructive

0

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

  • No destructive actions.
03 · HOW IT WORKS

Honeycomb in three steps.

  1. 01Your users connect HoneycombThey sign in to Honeycomb on Arc0 Connect, under your brand, and approve the access you ask for.
  2. 02You set the rulesReads run, and writes like “create slo” can wait for the user to approve.
  3. 03Any agent can actYour agent calls Honeycomb 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('honeycomb', {
  read: 'allow',
  write: 'ask',        // create_slo
  destructive: 'deny',  
})

# 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 Honeycomb connects.

Honeycomb runs its own MCP server behind OAuth. Arc0 registers the client, users approve access on Arc0 Connect, and your agent reaches Honeycomb through the same endpoint, policies and audit log as every other app.

The same Honeycomb 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
OAuth 2.0 (MCP)
CREDENTIALS
Per-tenant encrypted vault
MODEL SEES
Results only, never credentials
AUDIT LOG
Every call, on every plan
05 · WORKS WITH

Use Honeycomb from any agent.

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

More developer tools apps.

GitHub896 actionsSupabase129 actions0CodeKit8 actionsAbly25 actionsAbstract3 actionsAddressfinder17 actionsAddresszen2 actionsAlchemy36 actions
ALL DEVELOPER TOOLS APPS →
07 · FAQ

Honeycomb and Arc0, answered.

Q01

Can I use Honeycomb with Claude, ChatGPT or Cursor?

Yes. Connect Honeycomb 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 Honeycomb actions you allow.

Q02

How do users connect Honeycomb?

Honeycomb runs its own MCP server behind OAuth. Arc0 registers the client, users approve access on Arc0 Connect, and your agent reaches Honeycomb through the same endpoint, policies and audit log as every other app.

Q03

Which Honeycomb actions can my agent take?

33 in total: 16 read, 17 write and 0 destructive, such as “create slo”. Your policies decide which of them each agent may call.

Q04

Can I make my agent read-only in Honeycomb?

Yes. Allow read actions and deny writes in the Honeycomb policy. Your agent can still look things up, and any write it attempts is blocked and logged.

Q05

Can my own backend call Honeycomb too?

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

Get started

Plug Honeycomb into your agent.

Your users connect Honeycomb once, under your brand. Your agent gets 33 actions behind your policies, with every call on the record.

Connect HoneycombBrowse all apps →

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