feat: add Firecrawl web search provider

[flamerged]

Target: danny-avila/LibreChat:dev
Branch: feat/firecrawl-web-search
Commit: 251a9be176eb276f6d3d6401541011f0038f35fe

Dependency PR:

• Depends on agents PR: danny-avila/agents#192
• Dependency branch: feature/firecrawl-search
• Dependency target: main
• Dependency commit: f6ae68013d93310729444a2b2d99d1040f896366
• Merge/publish order: merge/publish the agents PR first, then bump @librechat/agents here before final merge if required by maintainers.

Why This Matters

Firecrawl gives LibreChat an AI-oriented web search option that can combine search discovery and content extraction through one provider. That is useful for teams that do not want to run a separate metasearch provider plus a separate scraper just to produce model-ready web context.

This does not replace SearXNG's privacy/metasearch niche. SearXNG remains valuable when users want a self-hosted metasearch engine aggregating many upstream search services. Firecrawl is valuable when the deployment wants one API focused on finding sources and returning clean, scrape-ready page content for agents.

References:

• Firecrawl Search API: https://docs.firecrawl.dev/api-reference/endpoint/search
• SearXNG overview: https://docs.searxng.org/

Summary

Adds Firecrawl as a selectable web search provider in LibreChat config/schema/auth/UI, including:

• Firecrawl URL/version/key fields through auth submission
• schema and runtime auth handling
• English locale keys and examples
• duplicate Firecrawl provider/scraper field handling
• focused backend/frontend/schema tests

Tests

• post-rebase branch status: clean, feat/firecrawl-web-search...origin/dev [ahead 1]
• packages/api src/web/web.spec.ts: pass, 54 tests
• client ApiKeyDialog.test.tsx: pass, 9 tests
• packages/data-provider src/config.spec.ts: pass, 119 tests
• packages/data-schemas src/app/web.spec.ts: pass, 15 tests
• JSON/YAML parse checks: pass
• git diff --check: pass
• git diff --check HEAD^ HEAD: pass

Locale/docs Notes

LibreChat policy says feature PRs update only client/src/locales/en/translation.json; non-English app locale files are automated externally and fall back to English. .env.example and librechat.example.yaml document the Firecrawl config; external feature docs belong in LibreChat-AI/librechat.ai if maintainers want a docs PR.

Residual Risks

Runtime end-to-end Firecrawl search should be verified after the agents package publish/bump and with FIRECRAWL_API_KEY configured.

[danny-avila]

firecrawl is already supported

[flamerged]

firecrawl is already supported

You are right that Firecrawl is already supported today as a scraper. I think this PR is a different integration point: it adds Firecrawl as a search provider option.

I checked the current dev branch, and LibreChat has the same provider split:

• SearchProviders includes SERPER, SEARXNG, and TAVILY
• ScraperProviders includes FIRECRAWL, SERPER, and TAVILY
• the API key dialog renders Firecrawl under the Scraper dropdown, not under Search Provider

So today a deployment can use scraperProvider: "firecrawl", but cannot select searchProvider: "firecrawl". This PR is intended to wire Firecrawl's /search endpoint through LibreChat config/schema/auth/UI, paired with danny-avila/agents#192 for the runtime search-provider implementation.

If the preferred direction is not to expose Firecrawl as a search provider, that is totally fair; I just wanted to clarify that the existing Firecrawl support is scraper support, not search-provider support.

@danny-avila

[danny-avila]

My mistake, I could've sworn we had firecrawl as a search provider since the firecrawl team itself has contributed code here. Will review as soon as I can