Home

HOW IT WORKS / THE BUG-TO-FIX LOOP

From first install.
To your next fix.The context goes with you.

One useful issue.
For you and your coding agent.

Add capture. See what broke. Give your agent the evidence. Review the proposed fix and watch the next release.

No credit card. Guided setup keeps your place.

BUGTAPE / THE BUG-TO-FIX LOOP01 / 05
BugTape / issue evidencecheckout · example
A BugTape issue with a failed checkout request, session replay and captured events
01 — Capture the whole story.
AGENT / MCPPREVIEW

you Why did checkout fail?

↳ get_repro_context

POST /api/checkout → 500
Replay · request · console · stack

↳ get_bug_events

One report. The context to start.

Illustrative workflow · synthetic data
One report. The context your agent needs.Watch the film

Product tour: real UI captures with synthetic data. Illustrative agent workflow.

LESS BACK AND FORTH

A bug report.
With a place to start.

A message that says “checkout is broken” leaves you guessing. BugTape brings the browser replay, failed request, console and stack into one issue. You and your agent start with the same evidence.

01 / MAKE IT YOURS

Choose your privacy.
Then add capture.

Create a project and capture key. Setup walks you through privacy, your SDK and a first report from your app. Its progress menu keeps your place.

  1. Choose what your app records.

    Show the report button or capture silently. Use standard replay, stricter text masking or error-only capture.

  2. Add the SDK for your runtime.

    Paste the generated snippet into your app. Keep server keys in your server environment.

  3. Send one report from your app.

    Open its evidence in BugTape. A console sample does not prove your app is installed.

Open guided setup →
<script type="module">
  import { init } from 'https://app.bugtape.ai/bugtape.mjs';
  init({ apiKey: 'bt_live_your_key_here' });
</script>
Use your project capture key.Read the browser guide ↗

Illustrative install examples use SDK defaults. Review privacy and use your own capture key in Setup.

All runtimes and install guides ↗

02 / SEE WHAT HAPPENED

The moment it broke.
The evidence beside it.

The browser SDK keeps a rolling buffer in memory. When it reports an uncaught error, an unhandled rejection or a failed request after a click, the issue includes the captured context.

  • Replay the page and clicks.

    The replay is rebuilt from recorded DOM events. It is not a screen video.

  • Read the request, console and stack.

    Inspect the failure beside the activity that led to it.

  • Check your privacy choices.

    Typed input values are hidden by default. Review real app screens before release.

Read the privacy checklist ↗

Servers send error context without browser replay. Check capture support.

BugTape issue showing the failed checkout request, event counts, agent handoff and session replay
Real console capture · synthetic checkout data

03 / HEAR FROM YOUR USERS

“This looks wrong.”
Now you can see why.

A wrong price or a confusing screen might not throw an error. The optional bug-report button lets someone explain the problem in their own words, with the captured context attached.

  • Show the button.

    People can add a title, description and severity. The Reported by users filter gives these reports their own view in Issues.

  • Capture silently.

    Automatic error capture still works. Your own “Report a problem” link can open the form with open().

Screenshots and voice notes in the form are not available yet.

Set up the report button ↗
BugTape report form with a title, description and severity choice
The browser report form · illustrative product capture

04 / FIND THE FAILURE

Repeated reports.
One clear issue.

Reports of the same failure stay together. Read the count, releases and platforms in one place. When your app sends user IDs, you can see who was affected.

  • Keep your inbox useful.

    Open an occurrence when you need its exact context. Repeats do not become a new pile of issues.

  • Bring the alert to your workflow.

    Connect GitHub, chat, email or a webhook. Setup checks actual delivery where the connection supports it.

Connect your workflow →
BugTape issue inbox with grouped checkout failures, severity and report counts
Real issue inbox · synthetic product data

05 / GIVE YOUR AGENT A START

The agent gets context.
You keep the decision.

Connect Claude Code, Cursor or Codex over MCP and approve scoped access. Ask your agent to read the issue, inspect its evidence and propose a fix.

  • Read the same evidence.

    Your agent can get the replay context, request, console and stack without a manual copy-and-paste.

  • Review the proposed changes.

    Your agent does the code work. You review the pull request and decide what ships. A linked PR does not prove that a bug is fixed.

  • Watch the next release.

    Keep track of new reports and recurrence after you ship.

Connect your coding agent ↗
Agent handoff status from Sent through Picked up to a proposed pull request
Real handoff UI · synthetic product data

EXAMPLE / MCP WORKFLOW

list_recommended_critical
  ↓ choose an issue
get_repro_context
  ↓ read captured evidence
ack_bug
  ↓ report progress / link your PR

Illustrative workflow. Your agent does the code work.

Tools, access and credit costs ↗

LESS TIME EXPLAINING BUGS.

More time
making things.

Free to start. Built for you and your agents.