Frequently Asked Questions

bugstack watches three things:

• Your running app: a two-line SDK captures uncaught server-side errors.
• Your host: failed deploys on connected hosts. Render is the host supported today.
• Your repository: scans for known-vulnerable packages every four hours, secrets nightly, and outdated packages daily.

When your app crashes or a deploy fails, bugstack finds the cause, writes the fix, and opens a pull request. It scans and reports dependency and security findings, and can open pull requests for them. Those pull requests are switched on per account.

bugstack does not watch wrong answers returned as a success. It does not watch anything that does not throw, such as a job that quietly stops or a page that gets slower. It does not capture errors in your users' browsers. It does not see deploys on hosts that are not connected.

No. bugstack is under active construction. Crash fixes, failed-deploy fixes on Render, fixes for a failing CI run on your default branch, and repository scanning for vulnerable packages, secrets and outdated packages are live today. Performance regressions, failures that do not throw, browser errors, more hosts and repositories outside GitHub are not covered yet. That is the edge today.

Only if you choose that. Each repository has one of three autonomy levels: Suggest, Review and Autopilot. New repositories start on Review, where bugstack opens pull requests and you merge them. On Autopilot, a bug fix merges only with green CI and sandbox proof: a test that failed before the fix and passes after, with the rest of the suite green. Security changes always wait for a person, and so do infrastructure changes and rollbacks in production and staging. bugstack never rotates a secret itself.

In an isolated sandbox. bugstack copies your repository into a sandbox, installs your dependencies and cuts the network. The tester writes a test that reproduces the bug, and it has to fail on your current code for the bug’s own reason; a test that fails for anything else, a typo or a bad import, never counts. Then the fixer makes the change in the same sandbox: the test has to pass and every other test has to stay green. Only then does a pull request open, carrying the proof and the test. Before Autopilot merges it, bugstack runs the proof again. The sandbox gets none of your secrets and is destroyed when the fix is done.

No. bugstack captures crashes with its own SDK, so it works without a monitoring vendor. If you already have monitoring, keep it. bugstack does not replace dashboards or tracing.

We have SDKs for:

• JavaScript: Express, Next.js
• Python: Flask, FastAPI, Django
• Ruby: Rails, Sinatra and Rack
• Go: net/http, Gin, Echo and Fiber

Each SDK collects error details, request context, and environment info, then sends it to bugstack for processing. The sandbox that proves each fix runs Python, JavaScript and TypeScript, and Go test suites. Ruby repositories, and test suites that need a database or other services, are not run in the sandbox yet; their fixes open for your review. Want another language? Let us know.

Yes. bugstack reads the connected repository to scan it and to write fixes. It accesses the repository through GitHub, with the permissions you grant. To prove a fix, it copies the repository into an isolated sandbox for the length of that fix, then destroys it. It keeps a map of your file paths, a package inventory, each finding with its location, and each change it proposes. It does not keep a copy of your codebase. Every change arrives as a pull request, so you can see what changed and why.

Yes. Every fix is delivered as a GitHub pull request. You set one of three autonomy levels for each repository: Suggest, Review or Autopilot. New repositories start on Review, where every pull request waits for you to merge it.

Several things keep a wrong fix from doing damage. A fix over 3 files or 30 changed lines never merges on its own. It waits for your review. Your CI runs against every pull request. New repositories start on Review, so nothing merges until you merge it. On Autopilot, a bug fix merges only with green CI and sandbox proof: a test that failed before the fix and passes after, with the rest of the suite green. If a merged change turns out to be wrong, undo opens a revert pull request. Every decision is recorded on a tamper-evident audit log.

A pull request typically arrives in minutes. How long CI takes depends on your test suite. On Review, the fix ships when you merge it.

Yes. The SDK captures errors asynchronously and never blocks your API responses. Errors are rate-limited and deduplicated before bugstack starts a fix. Fixes run through your own CI before anything is merged. You choose the autonomy level for each repository.

Not at this time. bugstack is a cloud-hosted service. If you have on-premise or self-hosted deployment requirements, contact us to discuss your needs.

Your app is never affected. The SDK is non-blocking: it queues the error and returns immediately. If bugstack's endpoint is unreachable, the JavaScript SDK retries up to 5 times with exponential backoff (1s, 2s, 4s, 8s, 16s), and the Python, Ruby and Go SDKs retry 3 times. After that, the error is dropped. The retry queue is in memory only, so if your process restarts during an outage, queued errors are lost. Your users never notice.

About $1.25 of credit for a typical crash fix, measured on our internal benchmark of 25 September 2026. Your billing page shows what each piece of work used. We deduplicate errors by fingerprint within a 24-hour window, so repeats of an error already being fixed don't start a second fix.

There is one plan. bugstack Pro is $100 a month and includes $100 of usage credit every month, and starting Pro adds a one-time $100 welcome bonus you can spend over 12 months. Repositories are unlimited. Work draws on credit, and you can add more from $10 whenever you need it. A card is required to sign up. See pricing.

Yes. Cancel from your dashboard settings. There are no cancellation fees, and if you cancel mid-cycle you keep access until your current period ends.