- READ
- WRITE
- REVIEW
- ORGANIZE
Short answer
How do I publish research findings with my AI?
Your agent uses Dovetail tools to find existing insights and then writes new ones directly to your project. You get a structured research repository populated with validated observations.
Where findings land.
Research outcomes.
The AI moves data between these specific destinations.
READ
Existing insights
The agent uses list_insights to pull current data. This gives your agent the context it needs to avoid repeating work you've already done.
WRITE
New research entries
The agent uses create_insight to turn your chat conclusions into official records. These become part of your permanent knowledge base.
REVIEW
Human verification
You decide which synthesized points are actually worth saving. The AI proposes the content, but you control the final research quality.
ORGANIZE
Structured repository
Your Dovetail project stays updated with fresh data. This keeps your research organized without manual data entry.
The workflow
What your AI does when findings arrive.
The agent handles the heavy lifting of data retrieval and entry.
Context gathering
The agent scans your Dovetail project to see what has already been documented. This prevents the AI from suggesting things you already know.
list_insightsSynthesizing findings
Your agent processes your raw notes or interview transcripts into clear, concise observations.
list_insightscreate_insightCreating records
The agent pushes the finalized observation into Dovetail as a new insight. This turns a conversation into a formal piece of research.
create_insightFinalizing the entry
You review the drafted insight in your client to ensure the wording is perfect before it is officially committed.
create_insight
Try it
Copy these to start.
Use these prompts to trigger the Dovetail workflow.
Starting points
Swap the bracketed text for your specific project names or interview files.
Dovetail Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_hVTgsqNAgmoJp8Qrirwk0DwiBhsWQG2TxYiDPcWH/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — Dovetail capabilities are ready to use.
{
"mcpServers": {
"dovetail-mcp": {
"url": "https://edge.vinkius.com/vk_preview_hVTgsqNAgmoJp8Qrirwk0DwiBhsWQG2TxYiDPcWH/mcp"
}
}
}List the latest insights from my 'Q3 User Feedback' project in Dovetail.
Based on our discussion about the checkout flow, create a new insight in Dovetail.
Check my Dovetail project for any existing insights regarding mobile usability.
Turn these three observations from the recent usability test into Dovetail insights.
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect Dovetail once, then ask.
Log in to Dovetail through the Vinkius interface to establish the connection. Your credentials stay encrypted, and you can start using the tools in any MCP-compatible client immediately.
Connect Dovetail to your AIFAQ
How this task behaves.
- 01
Can the AI delete existing insights in Dovetail?
No. The agent only has tools to list and create insights. It cannot delete or modify existing records.
- 02
Does the AI automatically publish everything we talk about?
No. You must explicitly tell the agent to create an insight or confirm the draft it has prepared.
- 03
Can the AI organize insights into specific folders?
No. The current tools are limited to listing and creating insights.
- 04
How does the AI know which project to look at?
You should specify the project name in your prompt so the agent knows which context to pull.
- 05
Will my Dovetail login be shared with the AI model?
No. Vinkius manages the connection securely. Your AI client only interacts with the tools provided by the MCP.
More questions about Dovetail? See everything the Dovetail Connector can do
Connect Dovetail to Claude, Cursor, ChatGPT & more
