agentleFS
KumailQaziKumailQazi/krmaazha-binance-mcp/.agents/skills/mcp-registry-syndication/SKILL.md

mcp-registry-syndication

Skill4 starsChanged 38 days agoMITView on GitHub

Multi-registry publishing, verification, and syndication guide for MCP servers across the Official MCP Registry (registry.modelcontextprotocol.io), Smithery.ai, Glama.ai, and PyPI. Covers server.json schema validation, GitHub Device Flow auth, PyPI ownership proof (mcp-name tag), and version lockstep.

  • Reads credentials

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If this marker is missing from the package on PyPI, the publishing command will fail with an authorization/ownership mismatch error.

    line 67
  2. Very specific:

    Follow the one-time device code prompt at https://github.com/login/device.

    line 74
  3. Very specific:

    Create smithery.yaml in repository root:

    line 93

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

MCP Registry Syndication & Distribution Skill

This skill governs the publication, lifecycle maintenance, and cross-directory syndication of Model Context Protocol (MCP) servers across the global ecosystem.


1. Official MCP Registry (registry.modelcontextprotocol.io)

The Official Model Context Protocol Registry is maintained by the MCP steering group and provides standardized discovery for MCP clients (Claude Desktop, Cursor, Zed, Sourcegraph, etc.).

1.1 Manifest Specification (server.json)

The registry mandates adherence to the official schema: https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json.

{
  "$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json",
  "name": "io.github.<Owner>/<repo>",
  "title": "Human-Readable Title",
  "description": "Concise summary under 120 characters.",
  "version": "1.2.2",
  "websiteUrl": "https://github.com/<Owner>/<repo>",
  "packages": [
    {
      "registryType": "pypi",
      "identifier": "<package-name>",
      "version": "1.2.2",
      "transport": {
        "type": "stdio"
      },
      "environmentVariables": [
        {
          "name": "BINANCE_ENV",
          "description": "testnet or mainnet",
          "isRequired": true,
          "isSecret": false
        },
        {
          "name": "BINANCE_READ_API_KEY",
          "description": "Read-only API key",
          "isRequired": true,
          "isSecret": true
        }
      ]
    }
  ]
}

1.2 Namespace & Ownership Requirements

  1. Reverse-DNS Format:
    • Format: io.github.<Owner>/<repo>
    • Example: io.github.KumailQazi/krmaazha-binance-mcp
  2. PyPI Ownership Attestation:
    • The official registry crawler fetches the published PyPI package README.
    • It requires the HTML comment / subscript tag:
      <sub>mcp-name: io.github.<Owner>/<repo></sub>
    • If this marker is missing from the package on PyPI, the publishing command will fail with an authorization/ownership mismatch error.

1.3 Publishing Workflow via mcp-publisher

  1. Authentication (GitHub Device Flow):
    npx -y mcp-publisher login github
    Follow the one-time device code prompt at https://github.com/login/device.
  2. Validation & Dry Run:
    npx -y mcp-publisher publish --dry-run
  3. Live Publishing:
    npx -y mcp-publisher publish
  4. Verification: Inspect active registry entry: https://registry.modelcontextprotocol.io/v0.1/servers?search=io.github.<Owner>/<repo>

2. Directory Syndication Formats

2.1 Smithery.ai (smithery.yaml)

Smithery enables one-click CLI and web deployment (npx -y @smithery/cli run ...). Create smithery.yaml in repository root:

startCommand:
  type: stdio
  configSchema:
    type: object
    properties:
      BINANCE_ENV:
        type: string
        default: "testnet"
        description: "Binance environment (testnet or mainnet)"
      BINANCE_READ_API_KEY:
        type: string
        description: "Read-only Binance API key"
      BINANCE_READ_API_SECRET:
        type: string
        description: "Read-only Binance API secret"
    required:
      - BINANCE_ENV
      - BINANCE_READ_API_KEY
      - BINANCE_READ_API_SECRET
  commandFunction:
    bash: "uvx krmaazha-binance-mcp"
    powershell: "uvx krmaazha-binance-mcp"

2.2 Glama.ai (glama.json)

Glama provides online inspection, schema visualization, and directory ranking. Create glama.json in repository root:

{
  "$schema": "https://glama.ai/mcp/schemas/server.json",
  "name": "krmaazha-binance-mcp",
  "description": "Security-first Binance MCP with physical confirmation gate and quant research.",
  "transport": "stdio",
  "command": "uvx",
  "args": ["krmaazha-binance-mcp"]
}

3. Version Lockstep Anti-Drift Checklist

Before any release:

  1. pyproject.toml version matches target release (e.g. 1.2.2).
  2. server.json version matches pyproject.toml.
  3. server.json package version matches pyproject.toml.
  4. CHANGELOG.md contains top versioned section ## [1.2.2] — YYYY-MM-DD.
  5. Run consistency gate:
    pytest tests/test_consistency.py

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.