Cyanheads Server

---

Overview

Fleet discovery for the cyanheads MCP ecosystem, built on a hosted fleet.json catalog of pre-computed tool and server embeddings. Search the catalog by natural-language query, or resolve a known tool or server name to its description, connection URL, and per-client install snippet. Runs as a stdio process, a local Streamable HTTP server, or the public hosted endpoint above.

Tools

Tool	Description
cyanheads_search_catalog	Search fleet tools and servers by natural-language query. Returns ranked matches with brief summaries and the owning server.
cyanheads_describe_entry	Return the description, connection URL, and per-client install snippet for a named tool or server.

Capability reference

cyanheads_search_catalog _tool

• query 1–500 characters; scope selects tools (default) or servers result granularity
• Optional category filter: research, government, public-data, utility
• limit 1–20 (default 5); results below the SIMILARITY_FLOOR (default 0.3) are dropped before the limit applies
• Every result carries score (cosine similarity, [0, 1]), comparable only within one response
• For scope tools, a servers roll-up (top 10 by best-matching tool, serversTotal for the full count) summarizes which servers matched
• Throws retryable catalog_empty while the catalog is still loading

---

cyanheads_describe_entry _tool

• name 1–64 characters; accepts a snake_case tool name or kebab-case server name, auto-detected or pinned via kind
• Tool lookups return the description and owning server; server lookups return version, npm package, GitHub URL, the full tool list, and per-client install snippets
• client filters snippets to one of claude-code, codex, cursor, gemini, streamable-http, curl; omit for every client
• Local (stdio, via npx) snippets are returned for every published server; remote (Streamable HTTP) snippets are added only when a hosted endpoint exists
• Discriminated on kind (tool | server) so callers branch on data, not string parsing
• Throws not_found (unknown name), ambiguous_kind (name matches both a tool and a server — pass kind to disambiguate), or retryable catalog_empty

Features

Built on @cyanheads/mcp-ts-core: stdio and Streamable HTTP transports, pluggable auth (none / jwt / oauth), swappable storage (in-memory, filesystem, Supabase, Cloudflare KV/R2/D1), structured logging with optional OpenTelemetry tracing.

Catalog-specific:

• Hourly background catalog refresh (CATALOG_REFRESH_SECONDS, default 3600) with an atomic index swap when generatedAt changes — no restart needed
• Query embeddings are computed at request time via @huggingface/transformers; document embeddings are pre-computed, L2-normalized, and Matryoshka-truncated, shipped inside fleet.json
• The embedding model is warmed up during startup, before OpenTelemetry's HTTP instrumentation patches fetch — avoids a cold-cache model-load failure under OTEL
• Self-describing: cyanheads_describe_entry resolves this server's own name and tools from a static fallback record, consulted only when the remote catalog doesn't carry an entry for it
• CATALOG_URL can point at any endpoint serving the same schema, to front a custom fleet

Agent-friendly output:

• Search responses echo the effective query, report the total match count before the limit, and include broadening guidance when nothing matches
• Discriminated result.kind (tool | server) on cyanheads_describe_entry lets callers branch on data, not string parsing
• Every search result carries a score field for trust calibration, plus a servers roll-up so agents can see which servers matched without a second call

Getting started

Public Hosted Instance

A public instance is available at https://cyanheads.caseyjhand.com/mcp — no installation required. Point any MCP client at it via Streamable HTTP:

{
  "mcpServers": {
    "cyanheads-mcp-server": {
      "type": "streamable-http",
      "url": "https://cyanheads.caseyjhand.com/mcp"
    }
  }
}

For Claude Code:

claude mcp add --transport http cyanheads https://cyanheads.caseyjhand.com/mcp

Self-Hosted / Local

Add the following to your MCP client configuration file.

{
  "mcpServers": {
    "cyanheads-mcp-server": {
      "type": "stdio",
      "command": "bunx",
      "args": ["@cyanheads/cyanheads-mcp-server@latest"],
      "env": {
        "MCP_TRANSPORT_TYPE": "stdio",
        "MCP_LOG_LEVEL": "info"
      }
    }
  }
}

Or with npx (no Bun required):

{
  "mcpServers": {
    "cyanheads-mcp-server": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@cyanheads/cyanheads-mcp-server@latest"],
      "env": {
        "MCP_TRANSPORT_TYPE": "stdio",
        "MCP_LOG_LEVEL": "info"
      }
    }
  }
}

Or with Docker:

{
  "mcpServers": {
    "cyanheads-mcp-server": {
      "type": "stdio",
      "command": "docker",
      "args": ["run", "-i", "--rm", "-e", "MCP_TRANSPORT_TYPE=stdio", "ghcr.io/cyanheads/cyanheads-mcp-server:latest"]
    }
  }
}

For Streamable HTTP, set the transport and start the server:

MCP_TRANSPORT_TYPE=http MCP_HTTP_PORT=3010 bun run start:http
# Server listens at http://localhost:3010/mcp

Prerequisites

• Bun v1.4.0 or higher (or Node.js v24+).

Installation

1. Clone the repository:

git clone https://github.com/cyanheads/cyanheads-mcp-server.git

2. Navigate into the directory:

cd cyanheads-mcp-server

3. Install dependencies:

bun install

4. Configure environment:

cp .env.example .env
# edit .env to override any defaults

Configuration

All configuration is validated at startup via Zod schemas in src/config/server-config.ts. Every variable has a sensible default — out of the box, the server points at the canonical cyanheads fleet.

Variable	Description	Default
MCP_TRANSPORT_TYPE	Transport: stdio or http	stdio
MCP_HTTP_PORT	HTTP server port	3010
MCP_HTTP_HOST	HTTP server bind host	127.0.0.1
MCP_HTTP_ENDPOINT_PATH	HTTP endpoint path where the MCP server is mounted	/mcp
MCP_AUTH_MODE	Authentication: none, jwt, or oauth	none
MCP_SESSION_MODE	HTTP session posture: auto, stateful, or stateless. src/index.ts declares stateless; set this only to override it.	stateless
MCP_LOG_LEVEL	Log level (debug, info, warning, error, etc.)	info
CATALOG_URL	Remote fleet.json endpoint (schema v2 with baked embeddings). Must be an absolute URL. Override to front your own fleet.	https://caseyjhand.com/fleet.json
CATALOG_FETCH_TIMEOUT_MS	Per-request timeout for fleet.json fetches in ms. Must be > 0.	10000
CATALOG_REFRESH_SECONDS	Background poll interval for fleet.json refresh. 0 disables; otherwise must be > 0.	3600
EMBEDDING_MODEL_ID	Hugging Face model id for query embedding. Must match fleet.json.embeddingModel.	Snowflake/snowflake-arctic-embed-m-v1.5
SIMILARITY_FLOOR	Cosine similarity cutoff for cyanheads_search_catalog results. Must be within [0, 1].	0.3
OTEL_ENABLED	Enable OpenTelemetry	false

See .env.example for the full list of optional overrides.

Running the server

Local development

• Build and run the production version:

  # One-time build
  bun run rebuild
  
  # Run the built server
  bun run start:http
  # or
  bun run start:stdio
  

• Run checks and tests:

  bun run devcheck  # Lints, formats, type-checks, runs MCP and packaging linters
  bun run test      # Runs the test suite
  

Docker

docker build -t cyanheads-mcp-server .
docker run --rm -p 3010:3010 cyanheads-mcp-server

The Dockerfile defaults to HTTP transport, stateless session mode, and logs to /var/log/cyanheads-mcp-server. OpenTelemetry peer dependencies are installed by default — build with --build-arg OTEL_ENABLED=false to omit them.

Project structure

Directory	Purpose
src/mcp-server/tools	Tool definitions (*.tool.ts). Two tools — cyanheads_search_catalog and cyanheads_describe_entry.
src/services/catalog	Catalog service — remote fleet.json provider with atomic-swap refresh, vector index, snippet builders, the self-description fallback record, and the query-time embedding runtime (@huggingface/transformers, behind an injectable interface for deterministic tests).
src/config	Server-specific environment variable parsing and validation with Zod.
tests/	Unit and integration tests, mirroring the src/ structure.
docs/	Design doc and schema reference.

Development guide

See CLAUDE.md for development guidelines and architectural rules. The short version:

• Handlers throw, framework catches — no try/catch in tool logic
• Use ctx.log for logging, ctx.state for storage
• Register new tools in the tools array passed to createApp() in src/index.ts
• Wrap external data: validate raw → normalize to domain type → return output schema; never fabricate missing fields

Contributing

Issues are welcome. Run checks and tests before submitting:

bun run devcheck
bun run test

License

Apache-2.0 — see LICENSE for details.