[21/36] Complete OC-070 protocol surface polish - #56
Conversation
1350cee to
865fb1c
Compare
1058eef to
774c211
Compare
865fb1c to
86ab2bc
Compare
774c211 to
0aa5e8d
Compare
86ab2bc to
72ae2dd
Compare
0aa5e8d to
a703e12
Compare
|
Verdict: request changes, one customer-facing misdetection bug plus a redaction gap.
Tests: solid happy paths (protocol identity through tree/CodeLens/tools, Postman script mapping, OpenAPI webhook/callback diagnostics). Gaps line up with the bugs: no test pastes a plain URL or arbitrary text through the detector, and redaction tests only cover bearer/ws-header/grpc-metadata, missing cli, awsv4 and the nested OAuth2 branches. Deps: none added. |
|
Re-review (whole-stack pass, per REVIEW_GUIDE.md) The Postman event-mapping claim from the #35 thread is verified: Prior findings still open:
New:
Verdict: changes-needed. The URL-misdetected-as-gRPC path is customer-facing and contradicts this PR's own README, and Postman imports silently activate disabled scripts and orphan collection/folder test scripts, both invisible until an imported suite behaves differently than in Postman. |
|
The substantive findings are fixed in
I did not extract Validation:
Stack integrity was checked before push. Updated PR #53 → PR #56, PR #56 → PR #54, PR #56 → PR #71, and PR #56 → PR #72 all simulate cleanly. GitHub reports PR #56 and immediate child PR #54 as MERGEABLE/CLEAN. |
|
Response to the second review: the actionable items are addressed on the current head (efd3226).
The review is correct that the focused event-scope regression covers oc070SurfacePolish, not postmanImporter; I am not claiming otherwise. I am retaining the two small attachImportMetadata helpers. Each is private to its importer and marks that importer's metadata boundary; sharing a ten-line helper would couple otherwise independent importers without changing behavior or closing a defect. I validated the complete 37-PR composition with npm run build, npm run compile, all 539 tests, and all 47 demo validations passing. |
Summary
Verification
This is part 2 of 21 in a stack made with GitButler: