Repository navigation
Conversation
The bridge-side cap is already exercised behaviourally by test_process_request_rejects_oversize_body_before_running_app. The only contract unique to the source-text test was that main.py hands the cap to the bridge; the grep passed as long as any one of the three asgi.fetch call sites still had it. Drive all three (cached GET miss, uncacheable GET, POST) through Default.fetch and assert each passes max_body_bytes=100_000, the limit the user-facing "over 100 kB" message promises. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_019boFsNWgtPe2zZQxPFRRWp
…y bridge.fetch Review follow-ups: compare against MAX_SUBMITTED_BODY_BYTES rather than a literal (the contract is that the two caps agree), check the call count per subtest so a skipped call site cannot reuse the previous call's kwargs, and cover worker_asgi_bridge.fetch forwarding max_body_bytes to process_request, which no test (old or new) exercised. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_019boFsNWgtPe2zZQxPFRRWp
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012eisRQmQvg4TphwztxhQGJ
Clean merge: #19 touched figures, theme tokens and figure-content checks; this branch only touches the request-body-cap tests. Generated files were rebuilt with make build and check-social-cards; nothing drifted. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012eisRQmQvg4TphwztxhQGJ
adewale
changed the title
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.
Replace the source-text body-cap check with tests through the real request boundaries. Every cached-GET-miss, uncacheable-GET and POST call from
Default.fetchmust passMAX_SUBMITTED_BODY_BYTES;worker_asgi_bridge.fetchmust forward it and reject an oversize request before the app runs.The old assertion searched for a substring that occurred at three call sites, so losing one cap stayed green. Its bridge checks also pinned local spelling rather than request behavior. Keep the existing fake runtime harness and add no production logic or dependencies.
Current verification — 2026-11-01
test_every_bridge_call_caps_the_request_body_at_the_submit_limit. Every changed file was restored afterwards.This replaces the contradictory earlier description that called the tests unexecuted while also reporting executed controls. The earlier session's permission rejection is historical, not a current validation result.
No production or rendered-output changes. This PR does not retire unrelated migration/documentation tests or assert that every body-limit scenario is covered. The complete browser/build/example-verification pipeline was not rerun in this follow-up. Final-head CI remains the merge check.