Repository navigation
Conversation
Apply the release skill's consensus-verified 37-package changed-mode plan for review only; this commit does not publish packages. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: f77d7bd4-9f62-45e2-8805-5ead9000f286
Codecov Report✅ All modified and coverable lines are covered by tests. ❌ Your project check has failed because the head coverage (97.0%) is below the target coverage (100.0%). You can increase the head coverage or adjust the target coverage. Additional details and impacted files@@ Coverage Diff @@
## user/martintomka/20260812-release-agentic-skill #669 +/- ##
=================================================================================
- Coverage 100.0% 97.0% -3.0%
=================================================================================
Files 503 497 -6
Lines 57407 55074 -2333
=================================================================================
- Hits 57407 53474 -3933
- Misses 0 1600 +1600
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
Stack the review-only release preview on the latest release-skill branch, including the upstream exposed-dependency cascade fixes. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: f77d7bd4-9f62-45e2-8805-5ead9000f286
Apply the unified resolver's renamed-root exposure evidence: the ohno break propagates through templated_uri_macros_impl and templated_uri_macros as 0.5.0 breaking releases. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: f77d7bd4-9f62-45e2-8805-5ead9000f286
| @@ -1,5 +1,11 @@ | |||
| # Changelog | |||
|
|
|||
| ## [0.5.0] - 2026-08-13 | |||
There was a problem hiding this comment.
| ## [0.5.0] - 2026-08-13 | |
| ## [0.4.1] - 2026-08-13 |
even though the syn was uprgraded to version 3, form SemVer perspective there were no breaking changes. I believe when it to comes to macros, the skill must be little bit more careful when evaluating what is breaking change.
|
review this PR |
martintmk
left a comment
There was a problem hiding this comment.
Posted by an AI agent
Three release-artifact issues were found: two blocking changelog correctness problems and one non-blocking structure issue.
Coverage included the exact 104-file diff and all 37 paired package manifests, public APIs and docs, tests, performance, naming, telemetry, resilience, and consistency. No runtime or public API behavior changed; cargo metadata, targeted cargo check, and git diff --check passed.
|
|
||
| - ✨ Features | ||
|
|
||
| - add configurable fast instant retrieval ([#648](https://github.com/microsoft/oxidizer/648)) |
There was a problem hiding this comment.
Posted by an AI agent
Generated changelog references use invalid GitHub paths
Problem
All 52 PR references newly added across 23 changelogs omit the /pull/ segment; for example, this link targets https://github.com/microsoft/oxidizer/648, which returns 404 instead of opening PR #648.
Why this matters
Every source link in these generated release notes is broken, so consumers cannot trace the changes being published.
Suggested fix
Restore /pull/ in the release generator's changelog link template and regenerate the affected changelogs.
|
|
||
| - ✨ Features | ||
|
|
||
| - add REST transcoding for gRPC services ([#600](https://github.com/microsoft/oxidizer/600)) |
There was a problem hiding this comment.
Posted by an AI agent
Release notes attribute unrelated changes to these packages
Problem
This entry attributes REST transcoding from PR #600 to data_privacy_core, but that PR changed only this crate's generated README links. The same false attribution appears in data_privacy_macros_impl, which also credits PR #585 even though its only package change was manifest array formatting.
Why this matters
The published changelogs would claim functionality that these packages did not gain, misleading consumers about the contents of the release.
Suggested fix
Exclude generated README-only and non-semantic formatting changes from package attribution, then regenerate the affected changelogs.
|
|
||
| - export frozen readers and guarantee dense handles ([#657](https://github.com/microsoft/oxidizer/657)) | ||
|
|
||
| All notable changes to this crate are documented here. The format is based on |
There was a problem hiding this comment.
Posted by an AI agent · Non-blocking
Changelog preamble is nested inside a released section
Problem
The Keep a Changelog/Semantic Versioning preamble was moved below the new 0.1.1 entries, making it part of that released section. The same relocation occurs in http_path_template and rest_over_grpc.
Why this matters
The document structure now presents repository boilerplate as release content and makes the generated changelogs harder to interpret.
Suggested fix
Treat the preamble as file header text and regenerate these changelogs with it before the first version heading.
Martin Taillefer (geeknoid)
left a comment
There was a problem hiding this comment.
Early-feedback static review found one release-note correctness gap: the plan correctly raises the two implementation crates whose public APIs expose syn, but the generated changelogs do not clearly identify that syn 2 → 3 compatibility break. This is also the one concrete testing gap: the changelog path needs a root-only workspace-dependency fixture.
The release tuple mechanics otherwise hold. All 37 crate manifest versions match the root workspace requirements, top changelog headings, and Cargo.lock; the lockfile contains no external dependency delta. The PR's 37-release total resolves to 29 Cargo-breaking transitions and 8 patch transitions, with no missing or non-increasing version.
Generated crate READMEs were excluded under AGENTS.md. Existing comments about invalid PR links, unrelated changelog attribution, and preamble placement were not duplicated.
All five taxonomies were assessed (150 security, 162 correctness, 182 testing, 183 performance, 143 conformance). This was static-only: nothing was built, tested, benchmarked, profiled, run under Miri, covered, mutation-tested, fuzzed, executed against a live target, or posted.
| @@ -1,5 +1,28 @@ | |||
| # Changelog | |||
|
|
|||
| ## [0.11.0] - 2026-08-13 | |||
There was a problem hiding this comment.
The actual breaking dependency upgrade is absent from these release notes — Conformance · Medium · High confidence
This package is raised from 0.10.3 to 0.11.0 because its public API permits syn::error::{Error, Result} and the workspace moved from syn 2 to 3, but this entry contains neither the syn upgrade nor a breaking-change section. The same root cause is visible in thread_aware_macros_impl's entry, where PR #621 is categorized only as a task even though the PR description classifies the exposed syn transition as breaking. Because the dependency change occurred in the root workspace manifest rather than under this crate directory, path-scoped changelog history cannot reconstruct the significant consumer-facing change. This conflicts with C-RELNOTES, which requires breaking changes to be clearly identified.
Direction: pass classified workspace-dependency changes and their exposure evidence into changelog generation instead of deriving release notes solely from crate-local history; render the syn 3 transition as breaking for both implementation crates.
Done when: both release entries explicitly identify the syn 2 → 3 public-API compatibility break, and a release-script fixture proves that a root-only workspace dependency bump exposed through a package API appears in that package's breaking notes.
Important
This is a release simulation only generated by the release skill from PR #662. Do not merge this PR to publish packages. No packages have been published.
Purpose
Dogfood the changed-package workflow against the current Oxidizer history so reviewers can inspect classifications, dependency cascades, exact version targets, changelogs, and generated READMEs while the release skill is being refined.
Result
templated_uri_macros_implandtemplated_uri_macrosfrom0.4.1patch previews to0.5.0breaking previews.Note
Codecov reports 100% patch coverage: all modified and coverable lines are covered. Its project check compares the stacked base's path-filtered 100% result with a broader 97% workspace run, so that project-level failure is not actionable through tests in this source-free simulation.
Version plan
data_privacy_coredata_privacy_macros_implfetch_optionshttp_path_templateinternityohno_macrosthread_aware_macros_impldata_privacy_macrosohnorouterama_buildthread_aware_macroscachet_tierdata_privacyfetch_tlsrouterama_macrostemplated_uri_macros_implthread_awareanyspawnbytesbufcachet_memorycachet_servicerouteramatemplated_uri_macrostickuniflightanyspawn_azurebytesbuf_iocachetmultituderest_over_grpcseatbelttemplated_urihttp_extensionsfetch_hyperseatbelt_httpfetchfetch_azureClassification highlights
thread_aware_macros_implanddata_privacy_macros_impl: breaking because public APIs exposesyntypes across the workspace syn 2 to syn 3 upgrade.ohno_macros: breaking because previously accepted#[no_constructors]input now fails compilation.tickandinternity: nonbreaking public API additions missed by objective tooling.Validation
cargo-semver-checks 0.48.0completed for all 20 modified ordinary libraries.AI-generated release simulation for review.