For AI-assisted workflows

One agentic interface, two ways in

Access your Skimle project data from your favourite AI tools. Use the in-app Agentic Chat tab, or connect Claude Code, Cursor, Windsurf, and other tools over MCP — the same structured data, the same interface.

Two ways to work with your data

Easy to use as part of Skimle

Agentic Chat within Skimle

Chat with an AI agent directly inside Skimle. Ask questions, merge categories, explore patterns — the agent has full structured access to your project data, with every insight traceable back to source. Simply use the Chat panel available in every project.

For agentic AI users

External MCP connection

Connect Claude Code, Cursor, Windsurf, OpenClaw or any MCP-compatible tool to your Skimle project. The same structured data and tools available in Agentic Chat — now accessible from your own coding or writing environment.

Connect Skimle MCP in seconds

Follow simple instructions below if you are using Claude Code, Cowork or other agentic AI tool with MCP.

1

Enable external access in Skimle

For security, only projects you explicitly grant access to are available via MCP. Enable external access from the External Access page in your Skimle home.

  1. 1. Log in to Skimle and go to Home → External Access in the left sidebar.
  2. 2. Click Create API key, choose a project, access level (read or read & edit), and expiry. You are done!

Please note that for security reasons, only project owners can create API keys. Once you have an API key, all members can access it via MCP.

2

Add Skimle to Claude (or other AI tool)

Go to claude.ai/settings/connectors or select Settings... on your Claude app, click Add custom connector, type Skimle as the name, and paste the URL: https://skimle.com/api/mcp

Select Skimle from the list and press Connect. Claude will ask you to sign in to Skimle to authorise the connection.

Use the same https://skimle.com/api/mcp URL to connect other tools such as Cursor, Windsurf, OpenClaw or Claude Code with Skimle.

3

Say hello to Claude

Open a new conversation in Claude and ask it to list your Skimle projects and what it can do with them.

The read & write Skimle MCP tools are included in your subscription and do not use any credits. In the near future you will be able to upload files and run analysis directly through a parallel skimle-admin MCP that uses your credits.

What your AI tools can do

Tools covering every aspect of your qualitative project — reading documents, searching insights, and organising categories.

Read & utility tools

  • list_projects discover all projects you have access to
  • get_project project overview, statistics, and metadata
  • get_documents browse uploaded documents
  • read_document full document content in paginated chunks
  • get_insights extracted insights with source quotes
  • get_categories full category hierarchy
  • get_memos AI-generated analytical summaries
  • get_tags tag catalogue
  • get_notes analyst notes
  • get_metadata document metadata fields
  • search semantic search across documents and insights
  • find_entities named people, organisations, and places
  • analyze metadata distribution and temporal patterns

Write tools

  • manage_insights create, update, or delete insights
  • categorize_insights assign, move, or remove insights from categories
  • manage_categories create, update, or delete categories
  • manage_tags create, update, delete tags; add or remove from documents or insights
  • create_note save a researcher note to a document, insight, or category

Available on all plans

MCP access is included with every Skimle subscription. Credits are shared across analysis, Agentic Chat, and MCP — no separate tier or add-on required.

Ready to connect?

Start with the in-app Agentic Chat, or wire Skimle into your coding environment via MCP.

Learn more about Skimle's features