Stoory MCP Server

Hire a human specialist from your AI workspace — without leaving it.

AI can build most of your app. It can't get you GDPR-compliant, review your Stripe integration, design your landing page, or run your SEO. Stoory MCP lets your AI coding assistant find, brief, and hire the right human for that — without you ever opening Upwork, writing a job post, or explaining your problem from scratch.

What it does

Search Stoory's marketplace (400+ specializations: dev, legal, design, growth, operations) with a natural-language query.
Package a task: your AI assistant drafts the title, description, and budget, and picks only the files the specialist actually needs — never your whole repo.
Hand off safely: a private GitHub repo is created per task, shared only with the invited specialist.
Track progress: check status and pull results back into your AI workspace once the specialist is done.

You stay in control of the two decisions that matter — who to hire and what budget to approve. The agent does the rest.

See Tools below for the full list.

Quickstart

Requirements

Any AI client that supports MCP over Streamable HTTP (Claude Code, Claude Desktop, Claude.ai, Cursor, Windsurf, VS Code, Codex CLI, ...).
A free Stoory client account and API key (below).

1. Get a client API key

Sign up at stoory.io/login?step=register&u=Client_(user), then open MCP Server on the Stoory homepage to generate your client_api_key.

2. Add the server to your AI client

Pick your client below. The URL is always https://mcp.stoory.io/mcp, the header is always x-api-key, but where you put them differs per client.

Claude Code

claude mcp add --transport http stoory \
  https://mcp.stoory.io/mcp \
  --header "x-api-key: YOUR_API_KEY_HERE"

Or add directly to .mcp.json:

{
  "mcpServers": {
    "stoory": {
      "url": "https://mcp.stoory.io/mcp",
      "headers": { "x-api-key": "YOUR_API_KEY_HERE" }
    }
  }
}

Claude Desktop / Claude.ai

Settings → Connectors → Add custom connector. Paste https://mcp.stoory.io/mcp as the URL, open Request headers, add x-api-key with your key, then click Add.

(Request header auth is currently rolling out at Anthropic. If you don't see a "Request headers" option yet, use Claude Code or another client below in the meantime.)

Cursor

Add to .cursor/mcp.json (project) or ~/.cursor/mcp.json (global):

{
  "mcpServers": {
    "stoory": {
      "url": "https://mcp.stoory.io/mcp",
      "headers": { "x-api-key": "YOUR_API_KEY_HERE" }
    }
  }
}

Windsurf

Add to ~/.codeium/windsurf/mcp_config.json — note the field is serverUrl, not url:

{
  "mcpServers": {
    "stoory": {
      "serverUrl": "https://mcp.stoory.io/mcp",
      "headers": { "x-api-key": "YOUR_API_KEY_HERE" }
    }
  }
}

VS Code

Add to .vscode/mcp.json — note the root key is servers, not mcpServers, and each entry needs "type": "http":

{
  "servers": {
    "stoory": {
      "type": "http",
      "url": "https://mcp.stoory.io/mcp",
      "headers": { "x-api-key": "YOUR_API_KEY_HERE" }
    }
  }
}

Codex CLI

Add to ~/.codex/config.toml:

[mcp_servers.stoory]
url = "https://mcp.stoory.io/mcp"
http_headers = { "x-api-key" = "YOUR_API_KEY_HERE" }

3. Ask for help

"My AI assistant keeps breaking my Stripe integration. Find me someone who can fix it."

That's it — the assistant searches, drafts the task, shows you cost and files before anything is sent, and waits for your go-ahead.

Tools

Tool	What it does
`search_specialists`	Semantic search over Stoory's specialist marketplace
`create_task`	Drafts and sends a task to one or more chosen specialists (requires explicit user confirmation)
`get_task_status`	Checks whether a specialist has accepted / is working / has finished
`get_task_result`	Returns the link to the task's private result repo
`get_task_repo_files`	Reads the result files directly, so the assistant can review the work itself
`cancel_task`	Cancels a task before a specialist has accepted it

How it works

"I need a GDPR specialist"
        │
        ▼
AI searches the marketplace, you pick a specialist
        │
        ▼
AI drafts the task + selects only the needed files
        │
        ▼
You confirm cost and file list
        │
        ▼
Private GitHub workspace created, specialist invited
        │
        ▼
You pull the result back into your AI workspace

Security & privacy

Your private data doesn't need to become public just because you need human help.

Your assistant chooses which files to share — never the whole repo.
Every task with attachments gets its own private GitHub repo, visible only to the invited specialist.
Files are scanned for secrets (.env, private keys, API keys/tokens) before anything leaves your machine.
Nothing is shared, and no task is created, without your explicit confirmation.

Example

User: "I've built my SaaS with AI. I need a privacy policy
       and terms of service before I can launch."

AI: 5 relevant specialists found.

User: "Use #2, budget $200."

AI: Creates the task, writes the brief, selects the relevant
    project files, opens a private repo, and shares it with
    the specialist. Done.

Community & support

Marketplace & signup: stoory.io
Issues / bugs: GitHub Issues
MCP Registry entry: io.github.stoorygitserver/stoory-mcp

Are you a specialist?

Stoory MCP recommends specialists from Stoory's marketplace to clients — if you're a developer, designer, lawyer, marketer, or any other freelancer, a free profile makes you eligible to get invited. No commission, ever. Create your free profile at stoory.io/for-specialists.

Disclaimer

Stoory MCP does not replace your AI assistant and does not perform payments or hiring decisions on its own — it prepares the task and hands the decision to you. Stoory does not process payments: client and specialist settle payment for the work directly between themselves, entirely outside Stoory.

Stoory Hire Specialist