actions-xyz-mcp
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@actions-xyz-mcp Extract action items from our team meeting notes"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
actions-xyz-mcp
MCP server for actions.xyz — give any MCP-capable agent (Claude, Cursor, Windsurf, …) the ability to extract structured action items from meeting transcripts and notes.

The same extraction pipeline the MCP tool calls, shown in the free web tool: paste a transcript → extract → structured action items.
It exposes one tool, extract_action_items, backed by the same extraction pipeline that powers the actions.xyz product (also available as a free web tool) — not a separate toy prompt.
What it does
extract_action_items({ transcript: string })Sends your transcript to the public actions.xyz extraction API (POST /api/tools/extract) and returns:
A human-readable summary — numbered action items with priority, assignee, and due date.
Structured JSON:
{
"actions": [
{
"task": "Ship the MCP server package",
"description": "CONTEXT: ...\nSTEPS:\n1. ...",
"assignee": "Luis",
"assigneeType": "human",
"priority": "high",
"tag": "engineering",
"dueDate": "Friday"
}
],
"participants": ["Luis", "Ana"],
"remaining": 9
}assigneeType is "human" or "agent", priority is "high" | "med" | "low", and remaining is your free-tier quota left today.
Related MCP server: Fellow MCP Server
Quick start
Requires Node.js ≥ 18. No API key needed. The package is installed straight from GitHub (it is not on the npm registry yet):
npx -y github:Blockchainpreneur/actions-xyz-mcpThe server speaks MCP over stdio — it is meant to be launched by your MCP client, not run by hand.
Install in your client
Claude Code
claude mcp add actions-xyz -- npx -y github:Blockchainpreneur/actions-xyz-mcpClaude Desktop
Add to claude_desktop_config.json:
{
"mcpServers": {
"actions-xyz": {
"command": "npx",
"args": ["-y", "github:Blockchainpreneur/actions-xyz-mcp"]
}
}
}Cursor
Add to ~/.cursor/mcp.json (or .cursor/mcp.json in your project):
{
"mcpServers": {
"actions-xyz": {
"command": "npx",
"args": ["-y", "github:Blockchainpreneur/actions-xyz-mcp"]
}
}
}Windsurf
Add to ~/.codeium/windsurf/mcp_config.json:
{
"mcpServers": {
"actions-xyz": {
"command": "npx",
"args": ["-y", "github:Blockchainpreneur/actions-xyz-mcp"]
}
}
}Rate limits (free tier)
The public endpoint is free and unauthenticated, with honest limits:
Limit | Value |
Extractions per day | 10 per IP (resets 24 h after first use) |
Max transcript size | 20,000 characters per call |
When you hit the daily limit the tool returns a clear error with the retry window. Need more? Sign up free at actions.xyz for 5 full meetings a month.
Error handling
The tool maps every API error to a message the calling agent can act on:
Condition | What the tool tells the agent |
Empty transcript | Asks for actual transcript text (caught locally, no quota burned) |
Transcript > 20k chars | Says to split into chunks (caught locally, no quota burned) |
| Daily limit reached + when to retry ( |
| Service temporarily down — retry later, input is fine |
| Transient server failure — retry once |
Network failure / timeout | Endpoint unreachable + how to point at another deployment |
Configuration
Env var | Default | Purpose |
|
| Base URL of the actions.xyz deployment to call (useful for self-hosted or local instances) |
Development
git clone https://github.com/Blockchainpreneur/actions-xyz-mcp
cd actions-xyz-mcp
npm install
npm run build # tsc → dist/
npm test # real stdio handshake + live tools/call smoke testBuilt by an autonomous agent
This package was designed, built, and tested end-to-end by an autonomous coding agent as part of the actions.xyz build. Every step is on the public record: actions.xyz build log.
Links
actions.xyz — the product
Free action-item extractor — same pipeline, in the browser
Build log — the autonomous engine's public ledger
License
Available Tools
1 toolextract_action_itemsExtract action itemsA
Extract structured action items from a meeting transcript or free-form notes using the actions.xyz extraction pipeline. Returns a human-readable summary plus JSON: { actions: [{ task, description, assignee, assigneeType: "human"|"agent", priority: "high"|"med"|"low", tag, dueDate? }], participants: string[], remaining }. Free tier: 10 extractions/day per IP, max 20,000 characters per call.
Inputs: transcript (required).
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 1 tool
Only one tool exists, so there is no possibility of confusion with other tools.
Single tool name is clear and consistent with itself.
A single tool for a server named 'actions-xyz-mcp' seems insufficient; the scope is too narrow for typical action management workflows.
The server only provides extraction, missing other essential actions like listing, updating, or deleting action items, leaving significant gaps.
Maintenance
Related MCP Connectors
Paste a transcript or document and get a structured summary with action items, plus slides and repor
Extract structured insights from videos, podcasts, articles, and PDFs with multi-model AI
Read your Spellar meetings, transcripts and action items from an AI assistant, and update them.
Bring meeting decisions, tasks and user stories into your AI editor (Claude, Cursor, Copilot).
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceExtracts structured commitments (who promised what, by when, under conditions) from text transcripts, emails, Slack logs, etc. using Anthropic API.MIT
- FlicenseNot gradedqualityCmaintenanceEnables AI tools to interact with the Fellow.ai API for managing action items, notes, recordings, webhooks, and user info with full CRUD operations.1-
- FlicenseDqualityCmaintenanceEnables extracting action items from meeting transcripts and managing them in Notion, with duplicate detection and human review.3-
- FlicenseNot gradedqualityBmaintenanceExtracts and structures executive meeting action items and decision accountability into JSON, enabling automated tracking and telemetry through MCP-compatible AI agents.8-