You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
docs(devtools-extension): add screenshots and Learn tab section to README
Replace TODO screenshot placeholders with actual screenshots of the
Flow, Timeline, and Learn views. Add Learn section documenting the
canvas-based request lifecycle visualization.
Copy file name to clipboardExpand all lines: packages/devtools-extension/README.md
+12-5Lines changed: 12 additions & 5 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -4,8 +4,7 @@
4
4
5
5
Most auth debugging starts in the Network panel and stays there — copying tokens into jwt.io, cross-referencing timestamps, guessing which 400 was the CORS preflight and which was a bad grant. Ping DevTools replaces that with a single panel that captures both network traffic and SDK-level events, correlates them into flows, and runs an automated diagnosis engine that tells you _what went wrong and how to fix it_.
6
6
7
-
<!-- TODO: add screenshot — Flow view with the Flow Health error banner visible -->
8
-
<!--  -->
7
+

9
8
10
9
---
11
10
@@ -70,8 +69,7 @@ Ping DevTools gives you:
70
69
-**Flow-level structure** — the Flow view shows the authentication flow as a sequence of nodes with detail cards, not a flat list of URLs.
71
70
-**Playback** — step through the flow to see exactly what the SDK saw at each point.
Network events follow a parallel path: `network-observer.ts` uses `chrome.devtools.network.onRequestFinished` to capture HAR entries, filters them against auth URL patterns, and sends them to the same service worker.
@@ -218,6 +217,14 @@ A chronological table of all captured events. Each row shows timestamp, event ty
218
217
219
218
A visual representation of the authentication flow as a sequence of SDK nodes. The node rail draws coloured circles for each node with arrows connecting them, status and node-name labels, and a glow effect on the selected node. Selecting a node opens a detail card with contextual information — collectors for DaVinci, callbacks for Journey steps, phase and error data for OIDC — plus any causally linked network requests with expandable request/response bodies. The Flow Health banner appears above the rail when the diagnosis engine detects issues.
220
219
220
+
### Learn
221
+
222
+
A canvas-based visualization that maps the request lifecycle across four stages: **Browser**, **Server**, **SDK**, and **Form**. Each stage is drawn as a labelled card with animated connector arrows showing the direction and outcome of each hop — method labels on outgoing edges, status codes on responses. Error states are highlighted with red borders and status annotations (e.g. `X 400`), making it immediately clear where a request failed and how the SDK interpreted the result.
223
+
224
+
The Learn tab correlates network events with SDK state transitions to show the full round-trip: the browser sends a request, the server responds, the SDK processes the response into a node transition, and the form renders the result. When errors occur, you can see exactly which stage failed and how that failure propagated through the rest of the pipeline.
225
+
226
+

0 commit comments