What is a source URL? Definition for bug reports

A source URL is the exact page address where a finding was observed: the link a developer, designer, or AI agent can open to reproduce the state the reviewer saw.

It is the cheapest piece of metadata to capture and the most expensive to lose. A bug report missing the source URL forces the recipient to reverse-engineer where the screenshot came from, which often means messaging the reviewer; a report that includes it lets reproduction start with a click.

What the URL does and doesn't include

The URL covers route and query parameters. It does not cover authentication state, cookies, A/B test variant, feature flag, or any value held in JavaScript memory. For most bugs that is fine: the route gets the engineer 90% of the way. For state-dependent defects, note the extra context next to the URL: "logged in as a trial user, in the EU billing flow, with the new pricing flag on."

For AI-agent workflows the source URL is what makes the agent able to verify a fix. The agent can open the same URL, render the page, and check whether the regression is gone. Without a URL the verification step has nothing to act on.

Frequently asked questions

Why does the source URL matter on a bug report?

Without it, the engineer has to guess which page or which route produced the screenshot. With it, the first step of reproduction is a single click.

What about state that the URL doesn't capture?

Some state lives outside the URL: cookies, signed-in account, query parameters set by JavaScript. Note those next to the source URL when they are part of the repro.