feat(evals): add Claude Opus 5.5 and GPT 6 Sol (high) results

[benjamincanac]

Adds Claude Opus 5.5 (Claude Code) and GPT 6 Sol (high) (Codex) to public/agent-results.json with their per-eval rows. Existing experiments are untouched.

Both already pick up their icons through the claude and gpt prefixes in the evals page model map, so no code change needed.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
nuxt	Ready	Preview	Sep 22, 2026 8:16pm UTC

[coderabbitai]

Navigate logical layers of code changes, visualize relationships, and explore their blast radius.

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Advanced

Run ID: 6fdb6dd8-ac2e-4a5b-bfcb-2563f5b9e00f

📥 Commits

Reviewing files that changed from the base of the PR and between c0fb180 and 0728ace.

📒 Files selected for processing (1)

• public/agent-results.json

Included review availability: Your plan provides up to 2 included reviews per hour; 1 remains after this review.

---

📝 Walkthrough

Walkthrough

The benchmark export timestamp was updated. A new claude-opus-5.5 experiment was added with aggregate duration, cost, and pass-rate metrics. The results object now includes 31 evaluations from nuxt-000 through nuxt-ui-007. All evaluations passed on the first run except nuxt-020-fix-nuxt-module, which passed after one retry.

Priority: ⬇️ Low

Estimated code review effort: 2 (Simple) | ~10 minutes

Merge Risk: ⚪ Minimal · up to 0728a

The PR adds internally consistent benchmark results and presents no actionable merge-blocking risk.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 0…
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Title check	✅ Passed	The title accurately identifies the addition of Claude Opus 5.5 results. The available changes summary does not confirm the GPT 6 Sol results, but the title remains related to the changeset.
Description check	✅ Passed	The description directly relates to the changeset by describing additions to public/agent-results.json and the use of existing icon mappings.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[nuxt-com-bundle-report]

Production bundle

Comparing c0fb1805 with 93d21f87. Compressed sizes are calculated from the emitted production assets.

Metric	Base (Brotli)	PR (Brotli)	Δ Brotli	Δ gzip
Client JavaScript	2.54 MiB	2.54 MiB	+926 B (+0.0%)	+1.2 KiB (+0.0%)
Client CSS	30.3 KiB	31.4 KiB	+1.1 KiB (+3.6%)	+1.4 KiB (+3.6%)
Other client assets	339.3 KiB	339.3 KiB	—	—
Total client assets	2.90 MiB	2.90 MiB	+2.0 KiB (+0.1%)	+2.6 KiB (+0.1%)

Largest module increases

Module	Base (Brotli)	PR (Brotli)	Δ Brotli
/public/agent-results.json	12.3 KiB	13.2 KiB	+889 B (+7.1%)
/app/pages/docs/async-data-chunk-5.js	0 B	181 B	+181 B
/app/pages/blog/async-data-chunk-3.js	0 B	169 B	+169 B
/app/pages/docs/async-data-chunk-4.js	0 B	168 B	+168 B
/app/pages/docs/[version]/errors/async-data-chunk-8.js	0 B	153 B	+153 B
/app/pages/deploy/async-data-chunk-20.js	0 B	137 B	+137 B
/app/pages/blog/async-data-chunk-1.js	0 B	122 B	+122 B
/app/pages/deploy/async-data-chunk-19.js	0 B	121 B	+121 B
/app/pages/enterprise/agencies/async-data-chunk-7.js	0 B	120 B	+120 B
/app/pages/enterprise/async-data-chunk-6.js	0 B	118 B	+118 B

Module values come from Nuxt’s analyzer and are attribution estimates. This workflow is currently report-only.

Workflow run