Migrating from Deep Research

Migrate this codebase off Firecrawl's Deep Research API (/v1/deep-research) to the v2 Agent API (/v2/agent). Deep Research shuts down on Thursday, December 10, 2026, 15:00 UTC, and every /v1/deep-research request returns HTTP 410 after that.

1. Find every Deep Research call:
   - Node SDK: deepResearch, asyncDeepResearch, checkDeepResearchStatus
   - Python SDK: deep_research, async_deep_research, check_deep_research_status
   - These may be called on firecrawl.v1, FirecrawlAppV1, V1FirecrawlApp or an older FirecrawlApp
   - Raw HTTP: POST /v1/deep-research and GET /v1/deep-research/{id}
2. Upgrade the Firecrawl SDK to the latest version: npm install firecrawl@latest, or pip install -U firecrawl-py.
3. Rewrite each call on the v2 client, new Firecrawl({ apiKey }) in Node or Firecrawl(api_key=...) in Python:
   - deepResearch(query, params) -> agent({ prompt, schema, maxCredits }); Python deep_research(query, ...) -> agent(prompt=..., schema=..., max_credits=...). It starts the run and waits for it.
   - asyncDeepResearch -> startAgent; async_deep_research -> start_agent. Same arguments as agent; returns the job id.
   - checkDeepResearchStatus(id) -> getAgentStatus(id); check_deep_research_status(id) -> get_agent_status(id).
   - POST /v1/deep-research -> POST /v2/agent; GET /v1/deep-research/{id} -> GET /v2/agent/{id}.
   Request:
   - query (or topic), systemPrompt, analysisPrompt and jsonOptions.prompt -> merge into one prompt string.
   - jsonOptions.schema -> schema (a JSON schema, a Zod schema in Node or a Pydantic model in Python).
   - The finalAnalysis report -> add a required string field "report" to schema, described as a markdown report that answers the question.
   - Sources -> add a required "sources" array of { url, title } objects to schema.
   - maxUrls -> maxCredits (max_credits in Python). It caps what the run spends and defaults to 2500. There is no page limit.
   - maxDepth and timeLimit -> remove them. They have no equivalent.
   - onActivity / on_activity callbacks -> poll getAgentTrace(id) / get_agent_trace(id), or pass a webhook.
   - onSource / on_source callbacks -> read the "sources" field of the result.
   Response:
   - status keeps the same values: processing, completed, failed.
   - data.finalAnalysis -> data.report; data.json -> data; data.sources -> data.sources (from the schema field); data.activities -> the trace; error -> error.
   - A run that hits maxCredits stops with stopReason "credit_limit_reached" (stop_reason in Python).
4. Keep the program's behavior the same: same inputs, same outputs to its callers, same error handling.
5. Run the project's tests and fix anything that breaks.

Reference: https://docs.firecrawl.dev/migrating-from-deep-research.md