docs(cursor): agent mode support in guides

[smakosh]

Summary

Cursor now honors the custom OpenAI base URL for agent mode as well, not just plan mode. This updates every place that claimed the integration was plan-mode-only:

• apps/docs/content/guides/cursor.mdx — description, intro, the warn callout (now an info callout: plan + agent mode work; Tab autocomplete and inline edit stay on Cursor's backend), the test step (was suggesting testing autocomplete, which never routed), the What Works/Doesn't lists, and the troubleshooting entry.
• apps/ui/src/content/guides/cursor.md (llmgateway.io/guides/cursor) — description, intro caveat, feature table (Agent mode ✅), troubleshooting, and the closing benefits.
• Integration cards — apps/ui/.../integration-cards.tsx and the shared packages/shared/.../integration-guides-grid.tsx (which renders the apps/code guides grid): badge "Plan mode only" → "Plan + Agent mode", description updated.
• Blog fixes — kimi-k3.md ("Cursor (chat/plan mode only)" → "(plan and agent mode)") and kimi-k3-claude-code.md (Cursor section boundary note now only excludes inline edit + tab autocomplete).

Left untouched: the Q2 2026 roundup's "Cursor plan mode" line (historical record of what shipped at the time) and GitHub-Copilot-related "agent mode" mentions (different product).

Validation

• pnpm format clean
• turbo run build --filter=ui --filter=docs --filter=code passes

https://claude.ai/code/session_01AqXnpfo1aWzgCUw7hvqVPR

Summary by CodeRabbit

• Documentation
  • Updated Cursor integration guides to clarify support for both plan mode and agent mode through LLM Gateway.
  • Clarified that inline edit and tab autocomplete continue using Cursor’s backend.
  • Revised setup instructions, troubleshooting guidance, FAQs, and integration examples.
• UI Updates
  • Updated Cursor integration descriptions and badges to display “Plan + Agent mode” support.

[coderabbitai]

Walkthrough

Cursor documentation, integration metadata, and Kimi K3 content now state that Cursor plan and agent modes route through LLM Gateway, while inline edit and tab autocomplete remain on Cursor’s backend.

Changes

Cursor routing coverage

Layer / File(s)	Summary
Update Cursor routing guidance apps/docs/content/guides/cursor.mdx, apps/ui/src/content/guides/cursor.md	Guides now describe plan and agent mode routing, updated testing steps, and the continued Cursor-backend handling of autocomplete and inline edit.
Refresh integration metadata apps/ui/src/components/integrations/integration-cards.tsx, packages/shared/src/components/integration-guides-grid.tsx	Cursor descriptions and badges now use “plan and agent mode” and “Plan + Agent mode.”
Align Kimi K3 guidance apps/ui/src/content/blog/2026-07-19-kimi-k3-claude-code.md, apps/ui/src/content/blog/2026-07-19-kimi-k3.md	Kimi K3 content now documents Cursor plan and agent mode compatibility and the backend limitation for inline edit and autocomplete.

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

Possibly related PRs

• theopenco/llmgateway#3150: Updates related Cursor and Kimi K3 routing guidance in the same blog content.
• theopenco/llmgateway#3181: Changes integration metadata in shared registries.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title is concise and accurately reflects the main change: adding Agent mode support to Cursor documentation.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
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.

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch docs/cursor-agent-mode

---

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.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@apps/docs/content/guides/cursor.mdx`:
- Around line 90-91: Update the test success statement in
apps/docs/content/guides/cursor.mdx at lines 90-91 to refer specifically to plan
and agent requests. Update the FAQ in
apps/ui/src/content/blog/2026-07-19-kimi-k3-claude-code.md at lines 43-50 to
include agent mode while preserving the existing backend-only limitations.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro Plus

Run ID: bb8277a3-f8e2-4d42-b66c-a2f30e3df15a

📥 Commits

Reviewing files that changed from the base of the PR and between e901d89 and 04165c2.

📒 Files selected for processing (6)

• apps/docs/content/guides/cursor.mdx
• apps/ui/src/components/integrations/integration-cards.tsx
• apps/ui/src/content/blog/2026-07-19-kimi-k3-claude-code.md
• apps/ui/src/content/blog/2026-07-19-kimi-k3.md
• apps/ui/src/content/guides/cursor.md
• packages/shared/src/components/integration-guides-grid.tsx

[coderabbitai]

🎯 Functional Correctness | 🟡 Minor | ⚡ Quick win

Propagate the agent-mode routing scope to all dependent documentation.

The primary updates now advertise plan and agent routing, but adjacent statements still use broader or older chat/plan-only wording.

• apps/docs/content/guides/cursor.mdx#L90-L91: narrow the test success statement to plan and agent requests.
• apps/ui/src/content/blog/2026-07-19-kimi-k3-claude-code.md#L43-L50: update the FAQ to include agent mode while retaining the backend-only limitations.

📍 Affects 2 files

• apps/docs/content/guides/cursor.mdx#L90-L91 (this comment)
• apps/ui/src/content/blog/2026-07-19-kimi-k3-claude-code.md#L43-L50

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@apps/docs/content/guides/cursor.mdx` around lines 90 - 91, Update the test
success statement in apps/docs/content/guides/cursor.mdx at lines 90-91 to refer
specifically to plan and agent requests. Update the FAQ in
apps/ui/src/content/blog/2026-07-19-kimi-k3-claude-code.md at lines 43-50 to
include agent mode while preserving the existing backend-only limitations.