@alwaysmeticulous/debug-workspace 2.286.0 → 2.287.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/dist/templates/CLAUDE.md +27 -1
- package/package.json +2 -2
package/dist/templates/CLAUDE.md
CHANGED
|
@@ -243,12 +243,38 @@ replay), and drop into phase 4 only as needed.
|
|
|
243
243
|
9. **Logs index** -- grep `debug-data/logs-index/<role>-<replayId>.txt` by source
|
|
244
244
|
(`source=application`), type (`type=warn`), keyword, or virtual time (`vt=1234`).
|
|
245
245
|
|
|
246
|
+
### 3b. Check network match quality (when diffs involve partial matches)
|
|
247
|
+
|
|
248
|
+
When the replay shows "Low accuracy: Network issues", or `network-log/` shows pollyReplay
|
|
249
|
+
entries with a `repair=` suffix (indicating session repair was needed to serve a response)
|
|
250
|
+
in the root cause time span (events preceding the first visual difference), this is a
|
|
251
|
+
high-priority check:
|
|
252
|
+
|
|
253
|
+
1. Grep `timeline.ndjson` for `pollyReplay` events near the first divergence point. Each
|
|
254
|
+
event's `data` contains both:
|
|
255
|
+
- `data.pollyRequest.request.url` -- the **actual request** made during replay.
|
|
256
|
+
- `data.matchedRequest.request.url` -- the **original recorded request** whose saved
|
|
257
|
+
response was served (null if no match was found).
|
|
258
|
+
2. Compare these URLs for each pollyReplay event. If they differ significantly (different
|
|
259
|
+
endpoint, different query parameters, different resource), the response served almost
|
|
260
|
+
certainly doesn't match what the application expected. This is a common root cause for
|
|
261
|
+
diffs and should be flagged.
|
|
262
|
+
|
|
263
|
+
For example, if the actual request was `GET /api/v1/items` but Meticulous served a saved
|
|
264
|
+
response originally recorded for `GET /api/v1/orders?page=1&search=...`, the
|
|
265
|
+
response schema is likely wrong for the requesting code, causing downstream UI differences.
|
|
266
|
+
|
|
267
|
+
When you identify such a mismatch, report it as the likely root cause and note that the diff
|
|
268
|
+
can be disregarded -- it stems from a network matching issue, not a real code regression.
|
|
269
|
+
|
|
246
270
|
### 4. Deeper dives (as needed)
|
|
247
271
|
|
|
248
272
|
10. **Screenshot timeline context** -- `debug-data/screenshot-timeline-context/` for the 30
|
|
249
273
|
events before and 10 after each screenshot.
|
|
250
274
|
11. **Network activity** -- grep `debug-data/network-log/` for endpoints, status codes,
|
|
251
|
-
domains.
|
|
275
|
+
domains. For any pollyReplay entries with a `repair=` suffix, cross-reference with
|
|
276
|
+
`timeline.ndjson` to compare the actual request URL against the matched recorded request
|
|
277
|
+
URL (see phase 3b above).
|
|
252
278
|
12. **Virtual time progression** -- `diff` on `debug-data/vt-progression/` files to find
|
|
253
279
|
where head and base diverge.
|
|
254
280
|
13. **Replay parameters** -- `debug-data/params-diffs/` for computed diffs, or
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@alwaysmeticulous/debug-workspace",
|
|
3
|
-
"version": "2.
|
|
3
|
+
"version": "2.287.0",
|
|
4
4
|
"description": "Shared debug workspace pipeline for investigating Meticulous diffs and replays",
|
|
5
5
|
"license": "ISC",
|
|
6
6
|
"main": "dist/index.js",
|
|
@@ -14,7 +14,7 @@
|
|
|
14
14
|
"prettier": "^3.5.3",
|
|
15
15
|
"@alwaysmeticulous/client": "2.286.0",
|
|
16
16
|
"@alwaysmeticulous/common": "2.283.1",
|
|
17
|
-
"@alwaysmeticulous/downloading-helpers": "2.
|
|
17
|
+
"@alwaysmeticulous/downloading-helpers": "2.287.0"
|
|
18
18
|
},
|
|
19
19
|
"devDependencies": {
|
|
20
20
|
"vitest": "^4.1.4"
|