Repository navigation
docs(lens): a refused Browser Run call says nothing about the engine - #288
Merged
Merged
Conversation
The engine seam only flips kitesurfParamLive true on response.ok, so a call the API refuses falls through to chromium-rest whether or not the selector was honoured, and the span records that label beside the failure. On the free plan the daily browser-minute ceiling does most of the refusing, which makes this the common case rather than the edge: a log window fills with chromium-rest while holding no evidence about Kitesurf either way. Measured today while trying to answer whether /lens runs on Kitesurf. Filtering Workers Logs on exists(lens.render_engine) returned five lens.browser spans over 24h, every one chromium-rest with lens.outcome: browser_budget_spent, back to a 2026-08-07 theverge.com scan. Read the two attributes together, and read a window with no successful render as no data. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Deploying with
|
| Status | Name | Latest Commit | Updated (UTC) |
|---|---|---|---|
| ✅ Deployment successful! View logs |
aadhar-sh | 1432114 | Aug 09 2026, 12:44 AM |
Empty commit. The github-advanced-security check failed on 30d55ef inside GitHub's own autofix agent (CAPI 400, "The requested model is not supported") before it read the diff, and that run is not retriable through the API. This exists only to dispatch a fresh one. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
oddharsh
enabled auto-merge (squash)
August 9, 2026 00:43
oddharsh
added a commit
that referenced
this pull request
Aug 9, 2026
…rnoon (#296) Two releases shipped unlogged (Dyno at #295, the lens docs fix at #288), so /updates still names v179 as the running build while describing /perf, a path that now redirects. This stages the entry that covers both. It also records two findings from that work in CLAUDE.md, because neither is derivable from the tree and both cost real time: 26. `run_worker_first` caps at 100 rules and this repo sits at exactly 100. Adding two took it to 102 and wrangler refused to boot. Neither `npm run build` nor `wrangler deploy --dry-run` catches it, because parseStaticRouting runs at session creation; `npm run routes:check` does, because createTestHarness() starts a real Worker. That is a second reason the route oracle earns its keep. The note also says why deduping buys nothing (7 duplicates, wrangler counts the 100 unique) and to check the existing wildcards before adding a rule, which is how the Dyno change shipped at 100 instead of 102. 27. A red `github-advanced-security` check is usually GitHub's Copilot Autofix failing on its own infrastructure. Seen twice on 2026-08-08 across two unrelated PRs, both with CodeQL green. The signature is an empty output title plus an annotation pointing at `.github:<line>`, which is a directory and cannot have one. Read the sibling CodeQL check before touching code; it is not a required check and both PRs merged and shipped with it red. Co-authored-by: Aadharsh Pannirselvam <19518661+oddharsh@users.noreply.github.com> Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Chasing "does /lens actually run on Kitesurf" ran into a caveat worth writing down: an engine label on a non-200 is not evidence.
runBrowserActiononly flipskitesurfParamLivetrue onresponse.ok, so a call Cloudflare refuses falls through tochromium-restwhether or not the selector was honoured, and the span records that label beside the failure. On the free plan the daily browser-minute ceiling does most of the refusing, which makes this the common case rather than the edge.Measured 2026-08-08 in Workers Logs. Filtering on
exists(lens.render_engine)returned fivelens.browserspans over 24h, every onechromium-restwithlens.outcome: browser_budget_spent, reaching back to a 2026-08-07 scan of theverge.com on an earlier version. Read that as no data, not as a verdict for Chromium.Same family as the "a 200 was being read as proof Kitesurf served" note it sits under, so it goes directly after it. Docs only, no code touched.
🤖 Generated with Claude Code