docs: fix three docs still claiming CI runs on Blacksmith

[ivndev001]

What

Fixes the three documents #431 enumerates that still stated, in the present tense, that this repo's CI runs on Blacksmith — #430 removed Blacksmith from every executable surface, so these were false rather than merely dated.

Why

One of the three (08-bench-gate-verify.md) actively instructs hard-coding blacksmith-8vcpu-ubuntu-2404, citing "Per CI standard (CLAUDE.md)" — a citation that now points at a document saying the opposite. Anyone following that step would reintroduce exactly what #430 removed.

Changes

• docs/idea/11-testing-ci.md — CI section rewritten to the current facts: runner selection is a repository variable (vars.WURK_CI_RUNNER / vars.WURK_BENCH_RUNNER, ubuntu-latest fallback, fork PRs pinned unconditionally; release/deploy-demo/pages/dependabot and the spec-docs job stay on ubuntu-latest by design), Ruby via ruby/setup-ruby, one full suite run (no version matrix) per the CLAUDE.md CI standard, Redis-only service container, bench results to job summary + sticky PR comment on PRs matching bench.yml's paths filter. The "Test matrix (Ruby × Redis × Rails versions)" list item became "Test suite" and the release-gate "matrix" bullets were de-matrixed, because keeping them under rewritten present-tense text would reassert a second stale fact.
• docs/idea/06-performance.md — benchmark job runs where vars.WURK_BENCH_RUNNER points (fork pin, paths-filter condition), results to job summary + sticky comment vs the PR's base.
• docs/plans/2026/08/06/101-faster-than-sidekiq/08-bench-gate-verify.md — "Executed; the runner instruction is superseded" banner naming the stale runner instruction and citation; the step text itself is preserved unedited as an executed-plan record (bench.yml exists on main; status.yml marks the slice complete), per the issue's suggested scope.
• Left alone per the issue's scope: the changelog entry and the seed/roadmap mentions (history), and docs/idea/12-docs-site.md:27 (describes a VitePress/docs-site/ plan that shipped differently — design intent, classed harmless in docs: three documents still say CI runs on Blacksmith in the present tense (one instructs a Blacksmith SKU) #431; a follow-up could sweep it together with 00-seed.md:40's index row for this file).

Verification

• Diff is three markdown files only, no code.
• grep -in blacksmith over the two idea docs: 0 hits. Remaining hits in the plan file are the banner itself and the preserved superseded step.
• Runner facts re-verified against .github/workflows/ at head: 8 vars.WURK_CI_RUNNER + 1 vars.WURK_BENCH_RUNNER expression sites, all with fork pinning and ubuntu-latest fallback; ruby/setup-ruby confirmed; bench paths filter matched item-for-item against bench.yml:52-60; deploy-demo/release/pages/dependabot/spec-docs runs-on: ubuntu-latest confirmed.
• Two independent deep-review lenses (correctness + adversarial) ran on each head; every MINOR they raised was fixed in a follow-up commit and re-reviewed.
• bin/check was not run locally: this box has no Ruby toolchain outside the Docker verify harness, and a markdown-only diff cannot affect the Ruby suite. The PR's CI run is the gate, and the merge path is fail-closed on CI green.

Closes #431

[developerz-ai]

✅ Reviewed — nothing blocking

0 actionable comment(s) · diff-only (no code index)

_{🤖 developerz.ai — automated review, running on your model and your box.}

[coderabbitai]

Warning

Review limit reached

You’ve reached a temporary PR review limit under our Fair Usage Limits Policy.

Your current included review allowance is based on your included PR review attempts over the past 7 days.

Next review available in: 19 minutes

Limit details: You’ve used all 2 included reviews currently available. Your 56 included PR review attempts over the past 7 days set your current allowance at 2 reviews per hour.

You’re in a promotional period — use the checkbox below to run this review for free:

• Run review for free

_{On-demand reviews are free for the next 30 days. After that, they cost $0.25 per reviewed file.}

How can I continue?

Run this review now using the option above, or comment @coderabbitai review --use-credits.

You can also wait for the limit to reset, then comment @coderabbitai review or push new commits to the PR.

An organization admin can change what happens after included review limits in Billing.

How do review limits work?

CodeRabbit enforces per-developer PR review limits within each organization.

For paid Pro and Pro+ reviews, CodeRabbit uses a developer's included PR review attempts over the past 7 days to set the current hourly allowance. At typical activity levels, the full plan allowance applies. Higher sustained activity can lower the allowance until earlier attempts leave the 7-day window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Path: .coderabbit.yml

Review profile: CHILL

Plan: Pro

Run ID: c5028ec7-ae48-415a-bf76-2239216934d4

📥 Commits

Reviewing files that changed from the base of the PR and between 04348f2 and 2a9dcd6.

📒 Files selected for processing (3)

• docs/idea/06-performance.md
• docs/idea/11-testing-ci.md
• docs/plans/2026/08/06/101-faster-than-sidekiq/08-bench-gate-verify.md

---

_{Comment @coderabbitai help to get the list of available commands.}

[developerz-ai]

🤖 developerz.ai review — automated, running on your model and your box. What is this?

Review summary — 3 file(s), 0 finding(s).

Documentation-only PR fixing three docs that incorrectly stated CI runs on Blacksmith. Changes are accurate: updates runner references from Blacksmith to repository variables (vars.WURK_CI_RUNNER / vars.WURK_BENCH_RUNNER), updates Ruby setup to ruby/setup-ruby, and adds supersession banner to the plan file. No code changes. Clean PR.

Grounding — not available

This review ran without repo-wide grounding — code grounding was not available to the server that ran this review, so there was nothing for it to cite. It read the diff; treat a clean result here as weaker evidence than a grounded one.

Every finding must cite real code before it is posted. Grounding needs a CodeGraph index on the server running the review; until there is one, findings that cannot be cited are withheld rather than shown.

[developerz-ai]

🤖 developerz.ai review — automated, running on your model and your box. What is this?

Review summary — 3 file(s), 0 finding(s).

Documentation-only PR fixing three markdown files to reflect the current CI setup (repository variables instead of hard-coded Blacksmith runners). No code changes. The diff accurately updates docs to match the post-#430 CI configuration and adds appropriate superseded banners to executed plan records.

(1 of 5 review samples completed; findings below are the union of the completed samples.)

Incremental re-review requested; the full diff was reviewed (commit-scoped incremental is a follow-up).

[developerz-ai]

Looks good — docs updated to reflect current CI runner configuration, superseded banner added to the executed plan file. CI green, no risk signals. Ready to merge when convenient.

🤖 Posted by developerz.ai — the maintainer agent, not a human.

[developerz-ai]

🤖 developerz.ai review — automated, running on your model and your box. What is this?

Review summary — 3 file(s), 0 finding(s).

Documentation-only PR fixing three docs that incorrectly stated CI runs on Blacksmith. Changes are accurate to current CI configuration (repository variables for runner selection, ruby/setup-ruby, Redis-only service, no version matrix). The plan file adds a proper supersession banner. Diff is clean — only markdown files changed, no code.

Grounding — not available

This review ran without repo-wide grounding — code grounding was not available to the server that ran this review, so there was nothing for it to cite. It read the diff; treat a clean result here as weaker evidence than a grounded one.

Every finding must cite real code before it is posted. Grounding needs a CodeGraph index on the server running the review; until there is one, findings that cannot be cited are withheld rather than shown.

Incremental re-review requested; the full diff was reviewed (commit-scoped incremental is a follow-up).

[developerz-ai]

Looks ready — CI green, platform review approved, no risk signals. Ready for maintainer merge.

🤖 Posted by developerz.ai — the maintainer agent, not a human.

[ivndev001]

head_sha: 2a9dcd6
verdict: SAFE-TO-MERGE
lens_count: 2
lens_verdicts: correctness=SAFE,adversarial=SAFE
reviewer_run: attended-authorized-by-ivndev001-2026-08-21T09:12:12.551Z