docs(mcp): sync create_thread tool docs from common#8236

consensus-mcp.mdx

@@ -251,7 +251,7 @@ Contact Consensus for enterprise API access at [consensus.app/home/api](https://
 
 ## Search Tool
 
-The search tool searches Consensus for peer-reviewed research papers. ChatGPT Deep Research may call this tool multiple times with different queries and filters while building a research report.
+The search tool searches Consensus for peer-reviewed research papers. Use it for quick paper lists. For synthesis, comparisons, literature reviews, and other multi-step research questions, signed-in users with access can use `create_thread` when it is available. ChatGPT Deep Research may call `search` multiple times with different queries and filters while building a research report.
 
 ### Parameters
 
@@ -302,6 +302,70 @@ Each search returns:
 - "Recent research on large language model hallucination from top tier journals"
 - "Use Consensus Deep Research to compare evidence for different treatments for insomnia"
 
+## Research thread tools
+
+The `create_thread` tool dispatches the Consensus research agent for detailed, multi-step research questions. Use it for synthesis, comparison, contradiction, author-specific, or literature-review questions. For a simple list of papers, use `search` instead.
+
+<Note>
+  `create_thread` is only available to signed-in OAuth users who have been granted access. It does not appear for anonymous users or API-key callers.
+</Note>
+
+### Parameters
+
+| Parameter | Type | Required | Description |
+| --- | --- | --- | --- |
+| `input_message` | string | Yes | Your complete research brief. Include the precise question, scope, population, timeframe, comparisons, constraints, and desired output. |
+| `mode` | string | No | `pro` (default, ~20 papers, typically ~3 minutes) or `deep` (~50 papers, up to ~10 minutes, requires a paid plan). |
+| `filters` | object | No | Optional structured filters using the same keys as the `search` tool. |
+| `attachments` | object | No | Optional paper or collection attachments to scope the research. |
+
+### Response
+
+`create_thread` returns a dispatch confirmation while the research agent runs:
+
+<ResponseField name="thread_id" type="string">
+  The new thread identifier.
+</ResponseField>
+
+<ResponseField name="interaction_id" type="string">
+  The identifier for the dispatched research run.
+</ResponseField>
+
+<ResponseField name="title" type="string">
+  The thread title.
+</ResponseField>
+
+<ResponseField name="mode" type="string">
+  `pro` or `deep`.
+</ResponseField>
+
+<ResponseField name="url" type="string">
+  A direct link to the thread on Consensus, where the full analysis and visualizations render.
+</ResponseField>
+
+<ResponseField name="status" type="string">
+  `running` while the agent works.
+</ResponseField>
+
+<ResponseField name="poll_after_seconds" type="integer">
+  Minimum recommended wait (30 seconds) before checking back.
+</ResponseField>
+
+<ResponseField name="max_wait_seconds" type="integer">
+  Maximum expected wait for `pro` (~180 seconds) or `deep` (~600 seconds).
+</ResponseField>
+
+<Note>
+  Rich visuals such as the consensus meter, evidence-strength meters, results timeline, citation graph, and PRISMA diagram render only in the Consensus web app. They are not returned in the MCP response.
+</Note>
+
+### Example prompts
+
+- "What is the strongest evidence for and against remote work productivity?"
+- "Compare the long-term safety profiles of GLP-1 agonists for weight loss"
+- "Find and synthesize RCTs on CBT for anxiety since 2020, focusing on effect sizes and drop-out rates"
+- "What do we know about the relationship between gut microbiome and depression?"
+
 ## Troubleshooting
 
 <AccordionGroup>