Let an approved AI agent investigate

Give an authorized coding agent the backtrace, breadcrumbs, request context, notes, and history it needs through Telebugs MCP.

What can the agent actually see?

The useful debugging facts: projects, errors, reports, backtraces, breadcrumbs, tags, releases, notes, and the history around them. It can follow the evidence without asking you to ferry screenshots and copied JSON between tabs.

Can it change anything?

Only if you allow it. Read access is enough to investigate. Write access is separate and covers actions such as resolving, muting, assigning, merging, and adding notes. The tool can send retrieved context to its provider, so choose one whose data handling you accept.

Is this just copy-pasting a backtrace with extra steps?

No. MCP gives the coding agent structured access to Telebugs, so it can ask follow-up questions, open the exact report, and pull the next piece of context itself. Telebugs remains the source of truth. Follow the MCP setup guide to connect your tool.

What if I stop trusting a tool?

Revoke it from Connected MCP apps. Its access stops there, without hunting for a forgotten token or changing anything in your app.