Fix stale benchmark badge

[konard]

Fixes #52

Summary

• Replaced the stale README CI and Benchmark workflow-name badges with the active CI/CD Pipeline badge scoped to main via .github/workflows/release.yml.
• Added scripts/check-readme-badges.rs and wired it into the CI/CD lint job so markdown documentation cannot reintroduce deleted workflow badges, legacy workflow query links, missing workflow-file badges, or unscoped workflow badges.
• Added the requested case-study package under docs/case-studies/issue-52, including downloaded badge SVGs, Actions run metadata, relevant logs, the pre-fix README artifact, and root-cause analysis.

Root Cause

The old dedicated Benchmark workflow was removed during the workflow migration in commit 8def08d, but README badges still pointed to legacy workflow-name URLs. The downloaded SVG for the old Benchmark badge rendered as Benchmark - failing, while the current CI/CD Pipeline badge for main rendered as passing. The historical failing Benchmark logs show the old workflow running against stable Rust while old dependencies still required nightly-only features; that workflow is no longer active on main.

Reproduction

The saved pre-fix README is at docs/case-studies/issue-52/raw/readme-before-fix.md.

rustc scripts/check-readme-badges.rs -o /tmp/check-readme-badges
/tmp/check-readme-badges docs/case-studies/issue-52/raw/readme-before-fix.md

That command fails as expected; see docs/case-studies/issue-52/raw/pre-fix-badge-check.log.

Verification

• rustc scripts/check-readme-badges.rs -o /tmp/check-readme-badges
• /tmp/check-readme-badges passed on the fixed documentation.
• rust-script scripts/check-readme-badges.rs
• rust-script scripts/check-file-size.rs
• cargo fmt --all -- --check
• cargo clippy --all-targets --all-features
• cargo test --all-features --verbose
• cargo test --doc --verbose

Verification logs are saved in docs/case-studies/issue-52/raw/.

[konard]

Solution summary

Implemented and pushed the fix to PR 53: #53

Commit: 4ca0a6c0068b9e35b797340c9563a700344880a0

What changed:

• Replaced the stale README CI and Benchmark workflow-name badges with the active file-based CI/CD Pipeline badge scoped to main in README.md.
• Added scripts/check-readme-badges.rs to catch legacy workflow badge URLs, missing branch=main, and badge links to missing workflow files.
• Wired that check into CI at .github/workflows/release.yml.
• Added the requested investigation package under docs/case-studies/issue-52/analysis.md, including preserved badge SVGs, workflow logs, raw API data, timeline, root causes, and solution options.
• Updated PR title/body and marked PR 53 ready.

Root cause found: the README still referenced deleted legacy workflows by workflow name. The old Benchmark workflow’s last runs failed after it tried to install an old nightly toolchain while the checked-out repo selected stable, causing #![feature] may not be used on the stable release channel. The replacement follows GitHub’s current workflow status badge format: https://docs.github.com/en/actions/monitoring-and-troubleshooting-workflows/monitoring-workflows/adding-a-workflow-status-badge

Verification:

• Local reproducer confirmed the old README badge fails the new badge check.
• rust-script scripts/check-readme-badges.rs
• rust-script scripts/check-file-size.rs
• cargo fmt --all -- --check
• cargo clippy --all-targets --all-features
• cargo test --all-features --verbose
• cargo test --doc --verbose
• Fresh GitHub Actions run 24607776539 passed on the PR commit, including lint, coverage, Ubuntu/macOS/Windows tests, and package build: https://github.com/linksplatform/doublets-rs/actions/runs/24607776539

The working tree is clean, the PR is non-draft, and GitHub reports the merge state as clean.

Next useful follow-up:

1. Add a new dedicated benchmark workflow later if maintainers still want a separate Benchmark badge, using a current toolchain and file-based badge URL.

---

This summary was automatically extracted from the AI working session output.

[konard]

🤖 Solution Draft Log

This log file contains the complete execution trace of the AI solution draft process.

💰 Cost estimation:

• Model: gpt-5.4
• Provider: OpenAI
• Public pricing estimate: unknown
• Token usage: 330,038 input, 36,047 output, 12,785,920 cache read

📊 Context and tokens usage:

Total: (330.0K + 12.8M cached) input tokens, 36.0K output tokens

🤖 Models used:

• Tool: OpenAI Codex
• Requested: gpt-5.4
• Model: GPT-5.4 (gpt-5.4)

📎 Log file uploaded as Repository (38180KB)

• View complete solution draft log

---

Now working session is ended, feel free to review and add any feedback on the solution draft.

[konard]

✅ Ready to merge

This pull request is now ready to be merged:

• All CI checks have passed
• No merge conflicts
• No pending changes

---

Monitored by hive-mind with --auto-restart-until-mergeable flag