stock-mcp
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
stock-mcp
An MCP server that exposes DSA's 问股 (Agent Chat / ask-stock) capability as a cross-client tool.
What it does
Claude Desktop / Code / Cursor
|
| Streamable HTTP (MCP)
v
stock-mcp (this server)
|
| HTTP (REST)
v
DSA FastAPI API
|
/api/v1/agent/chat
/api/v1/agent/skillsask_stock — send a natural-language stock question, get the Agent Chat answer back.
list_agent_skills — list available strategy skills for
ask_stock.
Related MCP server: OpenFinClaw CLI
Prerequisites
Python 3.11+
DSA API server running and reachable (e.g.
python main.py --serve-onlyinside thedaily_stock_analysisrepo, defaulthttp://127.0.0.1:8000).
Install
cd stock-mcp
python -m venv .venv
source .venv/bin/activate
pip install -e ".[dev]"Or install from requirements.txt:
pip install -r requirements-dev.txtRun
# With defaults (DSA on 127.0.0.1:8000, no auth)
python -m stock_mcp
# Or via the installed console script
dsa-mcp-server
# With env overrides
DSA_API_BASE_URL=http://127.0.0.1:8000 DSA_API_AUTH_ENABLED=true DSA_API_PASSWORD=xxx python -m stock_mcpEnvironment variables
stock-mcp reads .env from the working directory automatically, or you can export variables directly.
Variable | Default | Description |
|
| DSA FastAPI base URL |
|
| MCP HTTP bind host |
|
| MCP HTTP bind port |
|
| MCP endpoint path |
|
| Whether DSA requires admin auth |
| (empty) | DSA admin password (required when auth enabled) |
|
| Timeout for short calls (skills, login) |
|
| Timeout for |
| (empty) | Comma-separated extra Host header values allowed by DNS rebinding protection. Localhost entries are always included. |
| (empty) | If set, also write logs to this file path. Parent directories are created automatically. |
|
| Rotate the log file when it reaches this size in bytes. |
|
| Number of rotated log files to keep. |
Logging to a file
By default logs go to stderr only. To also persist them to a file with rotation:
export DSA_MCP_LOG_FILE=/var/log/stock-mcp/server.log
export DSA_MCP_LOG_LEVEL=INFO
python -m stock_mcpLogs are then written to both stderr and DSA_MCP_LOG_FILE. Set DSA_MCP_LOG_MAX_BYTES=0 for a single ever-growing file.
Client configuration
Claude Desktop
Add to your claude_desktop_config.json (macOS: ~/Library/Application\ Support/Claude/claude_desktop_config.json):
{
"mcpServers": {
"dsa-ask-stock": {
"url": "http://127.0.0.1:8765/mcp",
"env": {
"DSA_API_BASE_URL": "http://127.0.0.1:8000",
"DSA_API_AUTH_ENABLED": "false"
}
}
}
}Claude Code
claude mcp add dsa-ask-stock http://127.0.0.1:8765/mcpCursor
In Settings → MCP, add a Streamable HTTP server pointing to http://127.0.0.1:8765/mcp.
Auth notes
If DSA is running with ADMIN_AUTH_ENABLED=true:
export DSA_API_AUTH_ENABLED=true
export DSA_API_PASSWORD="your-dsa-admin-password"
python -m stock_mcpThe server logs in once at startup via POST /api/v1/auth/login and reuses the session
cookie across tool calls.
Backend limitation
If DSA's AGENT_BACKEND is set to codex_app_server, ask_stock returns a clear error:
Codex requires the streaming Chat interface (/chat/stream), which is not exposed over
MCP in this tool. Switch AGENT_BACKEND to litellm (or auto) for MCP compatibility,
or use the DSA web Chat page directly.
Security
The server binds to 127.0.0.1 by default. When binding to a public interface (0.0.0.0), the
MCP SDK enforces DNS rebinding protection — only Host header values in the allowed list are
accepted (see DSA_MCP_ALLOWED_HOSTS). Keep the allowed list as narrow as your use case
requires.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
This server cannot be deployed
Maintenance
Related MCP Connectors
MCP server exposing the Backtest360 engine API as tools for AI agents.
MCP server for stocksense-ai documentation, generated by doc2mcp.
MCP server for OpenMM — exposes market data, account, trading, and strategy tools to AI agents
Trading MCP server for AI agents, with live market data, account reads and controlled execution.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceMCP server for A-share stock technical analysis and AI prediction, enabling LLM-based interaction to analyze stocks.345-
- FlicenseNot gradedqualityDmaintenanceEnables quant research, strategy generation, backtesting, and paper trading from natural language prompts, integrating with AI agents via an MCP server.24-

backtest360-mcpofficial
AlicenseAqualityCmaintenanceMCP server that exposes the Backtest360 backtesting engine API as tools, enabling AI agents to conversationally discover indicators, build and validate strategies, run backtests, and read results.1482 PyPIMIT- FlicenseNot gradedqualityDmaintenanceMCP server adapter that exposes A-share stock data tools, prompts, and resources via FastMCP, enabling querying of stocks, K-lines, financials, sectors, and market hot spots through natural language.-