feat: Add AI model configuration settings (Phase 3.4)

[Datakult0r]

Summary

Adds a settings page for configuring which AI models power each agent role in the research pipeline. Users can choose from three quality presets (Low-Cost FREE, Balanced $3.50, Premium $7) or customize individual agent models in advanced mode.

Backend (296 lines): New /api/settings routes with endpoints for listing models, presets, current config, updating settings, cost estimation, and reset.

Frontend (427 lines): New /settings page with preset cards, cost estimation display, advanced per-agent configuration, and cost breakdown visualization.

Review & Testing Checklist for Human

• Critical: Settings not integrated with research pipeline - The settings are stored but I did not verify they are consumed by researchService.js or agent_models.py. Check if changing presets actually affects which models are used during research.
• Security: Settings endpoints are unprotected - All /api/settings/* routes have no authentication. Decide if model configuration should require auth to prevent unauthorized cost increases.
• Persistence: In-memory storage only - Settings use let currentSettings = {...} and reset on server restart. For production, migrate to Supabase storage.
• Test UI flow - Navigate to /settings, select each preset, toggle advanced mode, change individual agent models, save, and verify the configuration persists (until server restart).

Recommended test plan:

1. Start backend and frontend locally
2. Navigate to /settings
3. Verify three preset cards render with correct costs
4. Click each preset and verify selection indicator appears
5. Toggle "Advanced" and verify per-agent dropdowns appear
6. Change a model, verify cost estimate updates
7. Click Save, refresh page, verify selection persisted

Notes

• This PR is based on the Session 9 branch, so PRs chore: repository cleanup - remove node_modules and resolve merge conflicts #10-18 should be merged first
• Cost estimates use hardcoded token approximations per agent (100K for PhD Student, 80K for Postdoc, etc.)
• Frontend gracefully falls back to demo data if API is unavailable

Link to Devin run: https://app.devin.ai/sessions/9573820f0f6d41dab1cff8a61c26c981
Requested by: @Datakult0r

[devin-ai-integration]

Original prompt from Datakult

https://github.com/Datakult0r/ai-grant-crawler-a2a-pro
index this

[vercel]

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

Project	Deployment	Review	Updated (UTC)
ai-grant-crawler-a2a-pro	Error		Jan 25, 2026 9:00pm

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR that start with 'DevinAI' or '@devin'.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment and CI monitoring

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

---

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 and usage tips.}