fix(web): include Codex in landing page runtimes

[waynesun09]

Summary

The landing page omits the supported Codex runtime. List “Claude Code · Pi · Codex” and use a two-column stats grid so the longer label does not leave uneven rows on desktop or narrow phones.

Related Issue

User-requested update; no linked issue.

Changes

• Add Codex and capitalize Pi in the Multi-runtime label.
• Keep the four hero stats in a responsive 2×2 grid.

Testing

• make lint passed with the change staged.
• git diff --check passed.
• Playwright verified widths of 320, 375, 390, 768, 900, 901, 1024, and 1440px; no horizontal overflow or clipped stats.
• Visually inspected desktop and mobile screenshots.
• No logic changes requiring unit tests.

Checklist

• PR title follows Conventional Commits.
• Commit signed off (DCO).

[fullsend-ai-review]

🤖 Finished Review · ✅ Success · Started 12:32 AM UTC · Completed 12:39 AM UTC

Commit: 21e30a7 · View workflow run →

Runtime: claude · Model: opus → claude-opus-4-6 · Effort: high · Cost: $1.45

[github-actions]

Site preview

Preview: https://c642d96a-site.fullsend-ai.workers.dev

Commit: 21e30a78369c75bd92548c812486f36694eeb1c3

[qodo-code-review]

PR Summary by Qodo

Include Codex in landing page runtime stats

🐞 Bug fix 🕐 Less than 5 minutes

AI Description

• Adds Codex and corrects Pi capitalization in the landing page runtime list.
• Uses a responsive two-column hero stats grid to preserve balanced rows.

Diagram

graph TD
  A["Landing Page"] -->|contains| B["Hero Stats"] -->|uses| C["2x2 Grid"] -->|displays| D["Claude Pi Codex"]

Loading

High-Level Assessment

The current approach is optimal for this static, fixed-size stats section. CSS Grid directly guarantees balanced 2×2 rows across viewport widths; retaining wrapping flexbox or adding breakpoint-specific rules would introduce less predictable layout behavior or unnecessary complexity.

Files changed (1) +2 / -2

Bug fix (1) +2 / -2

index.htmlShow Codex and balance the hero statistics grid +2/-2

Show Codex and balance the hero statistics grid

• Adds Codex to the multi-runtime label and capitalizes Pi consistently. Replaces the wrapping flex layout with a responsive two-column grid so all four hero statistics remain balanced.

web/public/index.html

[qodo-code-review]

Code Review by Qodo

🐞 Bugs (0) 📘 Rule violations (0) 📎 Requirement gaps (0)

Great, no issues found!

Qodo reviewed your code and found no material issues that require review

Tip of the day

💡 Did you know, you can commit Qodo's fix in one click with committable suggestions (GitHub & GitLab)

More tips ↗ | Customize Qodo ↗ | Qodo docs ↗

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[fullsend-ai-review]

Risk Assessment: low (1/5)

Details

Single-file HTML change with only 4 lines modified, no protected paths or dependencies touched, and low author contention; the modest fix/revert history on this file slightly elevates Tier 2 but not enough to move the composite above low.

[fullsend-ai-review]

Looks good to me

[fullsend-ai-retro]

🤖 Finished Retro · ✅ Success · Started 9:59 AM UTC · Completed 10:09 AM UTC

Commit: 21e30a7 · View workflow run →

Runtime: claude · Model: opus → claude-opus-4-6 · Effort: high · Cost: $3.21

[fullsend-ai-retro]

Retro: PR #7133 — fix(web): include Codex in landing page runtimes

This was a clean, well-executed workflow with no rework, no false positives, and correct verdicts from both agent and human reviewers.

Timeline: PR opened at 00:30 UTC → review agent approved at 00:39 UTC (9 min wall-clock, ~5 min agent runtime) → human approved and merged at 09:57 UTC.

What went well:

• The review agent correctly identified a trivial 4-line HTML/CSS change and approved with no findings.
• Sub-agent triage was reasonable: 3 mandatory/enabled sub-agents dispatched, 3 optional ones correctly skipped.
• The style-conventions sub-agent was particularly efficient, applying its early-exit rule in 13 seconds with only 255 output tokens.
• The correctness sub-agent did useful verification (confirmed Codex exists as a runtime, verified the 2x2 grid math against the 4 stat children).
• Human reviewer agreed with the agent's assessment — no delta between agent and human review.

Cost concern (already tracked): $1.45 and Opus-level effort for a 4-line cosmetic HTML change is on the high side. The multi-agent architecture has fixed overhead regardless of PR size. However, this improvement opportunity is already well-covered by existing open issues:

• fullsend#6891 — pre-flight scope heuristic to adjust effort/model for trivially-scoped PRs
• agents#497 — downgrade sub-agent models based on scope classification
• agents#879 — extend review fast-path to trivial human-authored PRs

This PR provides additional supporting evidence: a 4-line HTML-only change with no logic, no API changes, and no security surface still ran the full Opus orchestrator + 3 sub-agents pipeline.

Minor efficiency note: One wasted Bash call occurred when the orchestrator's shell command was blocked by the Tirith sandbox scanner and had to be retried with simpler syntax. The agent recovered correctly.

No new proposals filed — all identified improvement opportunities are already tracked in existing issues.