SDKs

One page per language, so you never scroll past a stack that is not yours. Every SDK shares one contract: batched shipping that never blocks your request path, errors reported with the stack and the log trail, and the same privacy rule — request bodies, headers, cookies and query strings are never captured.

Every SDK is open source under MIT, at github.com/InsightRecorder. Read what an agent running inside your application actually does before you install it — that is the point of shipping the source, and of keeping every one of them free of third-party runtime dependencies.

Every SDK takes a base URL (INSIGHT_BASE_URL): the address of your InsightRecorder deployment — the server that receives your data — not the application you are monitoring, which is where the SDK runs. Self-hosted, it is your own instance (e.g. https://insight.yourcompany.com); on the hosted service it is your InsightRecorder host.

In the browser

Capture the session that produced the bug. The agent is the front door; the adapters exist because React and Vue intercept component errors before the agent can see them.

On your servers

Ship logs and turn errors into bug reports. No third-party runtime dependency in any of them.

Looking for the endpoints instead? The full API reference is at /docs, and the source for every SDK is on GitHub.