Bug Recorder for AI agents
Bug Recorder captures application errors as they happen, feeding engineering teams a stream of failures to monitor and diagnose without digging through raw server logs. 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 Bug Recorder.
Log a caught exception
Send a captured error, with its message and context, into Bug Recorder for the team to triage.
Route errors from a live service
Forward exceptions thrown by a running service into Bug Recorder so failures don't go unnoticed.
Flag a recurring failure
Report a repeated error pattern so an engineer can decide whether it needs a fix.
1 Bug Recorder actions, graded by risk.
Every Bug Recorder action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.
read
0Look things up. Allowed by default.
- No read actions.
write
1Create and change things. Allow, or ask the user first.
- bug_recorder.send_errorSend Error
destructive
0Delete, cancel or archive. Ask first, or deny outright.
- No destructive actions.
Bug Recorder in three steps.
- 01Your users connect Bug RecorderThey add their Bug Recorder api key on Arc0 Connect, under your brand. It goes straight into the vault.
- 02You set the rulesReads run, and writes like “send Error” can wait for the user to approve.
- 03Any agent can actYour agent calls Bug Recorder through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
await arc0.policies.set('bug_recorder', { read: 'allow', write: 'ask', // send_error destructive: 'deny', }) # Claude Code: the same connection, one URL $ claude mcp add --transport http arc0 \ https://mcp.arc0.ai/u/u_8f2
How Bug Recorder connects.
Users add their Bug Recorder 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 Bug Recorder 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 Bug Recorder from any agent.
ClaudeChatGPTCursor
Codex
VS CodeOpenAI Agents SDKClaude Agent SDK
Vercel AI SDK
Mastra
LangGraph
Bug Recorder and Arc0, answered.
Can I use Bug Recorder with Claude, ChatGPT or Cursor?
Yes. Connect Bug Recorder 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 Bug Recorder actions you allow.
How do users connect Bug Recorder?
Users add their Bug Recorder 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 Bug Recorder actions can my agent take?
1 in total: 0 read, 1 write and 0 destructive, such as “send Error”. Your policies decide which of them each agent may call.
Can I make my agent read-only in Bug Recorder?
Yes. Allow read actions and deny writes in the Bug Recorder policy. Your agent can still look things up, and any write it attempts is blocked and logged.
Can my own backend call Bug Recorder too?
Yes. The same Bug Recorder connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.
Plug Bug Recorder into your agent.
Your users connect Bug Recorder once, under your brand. Your agent gets 1 actions behind your policies, with every call on the record.
Free to build · MCP + REST · Audit log on every plan