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., "@MCP Knowledge Bridge search for artificial intelligence trends"
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.
手写、可教学的 MCP 子集实现。v0.3.0 在 stdio 上支持 2025-11-25 协议协商，以及 tools / resources / prompts / ping 与初始化生命周期。体积刻意保持小，不是完整 MCP SDK，也不是生产网关。
python -m pip install -e ".[dev]"
python -m examples.demo_protocol
python -m mcp_bridge.serverstdio 传输：每行一条 UTF-8 JSON-RPC（非 Content-Length 分帧）。
mcp-knowledge-bridge{
"mcpServers": {
"knowledge-bridge": {
"command": "python",
"args": ["-m", "mcp_bridge.server"],
"cwd": "/absolute/path/to/mcp-knowledge-bridge"
}
}
}Related MCP server: Hoard
mcp_bridge.http_server 是带鉴权的 自定义 REST 伴生服务。
它不是 MCP Streamable HTTP，也不实现 MCP-over-HTTP / OAuth。
export MCP_BRIDGE_TOKEN="replace-with-a-secret"
mcp-knowledge-bridge-rest
curl -H "Authorization: Bearer replace-with-a-secret" http://127.0.0.1:8765/toolsPowerShell：
$env:MCP_BRIDGE_TOKEN = "replace-with-a-secret"已实现：initialize、notifications/initialized、ping、tools/list、tools/call、resources/list、resources/read、prompts/list、prompts/get。
未声称：sampling、roots、elicitation、subscriptions、Streamable HTTP 等。
python -m ruff check .
python -m pytest --cov=mcp_bridge --cov-report=term-missingMIT © Brian20040323
Echo payload for wiring tests.
Inputs: message (required).
Compare two knowledge topics and return a structured diff summary.
Inputs: left_id (required), right_id (required).
Fetch one knowledge entry by id.
Inputs: id (required).
BM25 search over local knowledge entries and markdown docs.
Inputs: query (required), top_k.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 4 tools
Each tool has a clearly distinct purpose: search, fetch by ID, compare two topics, and a debug echo. No overlap in functionality.
Three tools follow the 'kb_' prefix with verbs, but 'echo_debug' breaks the pattern by omitting the prefix. Mixed conventions reduce consistency.
Four tools is well within the typical 3-15 range and feels appropriate for a focused knowledge bridge server.
Only read and compare operations are present. Missing create, update, and delete for knowledge entries, which are essential for managing a knowledge base.
Persistent memory, knowledge graph, task management and more for AI agents via MCP.
Ask your own documents and get answers that cite the source passage. Workspace knowledge bases over MCP for Cursor, Claude and any MCP client.
A second brain for your AI agent. Shared memory for your team's AI agents, plus public and private knowledge bases — works with ChatGPT, Claude, Cursor and any MCP client.
Agent-native notes, tasks, dev-docs, vaults, sync & handoffs. MCP + OpenAPI dual surface.
