M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
A

AKShare One MCP

Community
An MCP server providing comprehensive China stock market data interfaces based on AKShare.
How to install ↓ Compare GitHub source repository ↗
Category
Other
Stars
★ 233 Popular
Transport
stdio (local process) · Streamable HTTP
Runtime
Node.js · Python
Credentials
No credential needed
License
MIT
Last commit
Tools
9
42FMRS · D

This server provides rich China A-share data interfaces including historical quotes, real-time data, news, and financial statements, suitable for users needing Chinese stock data. However, it relies on third-party data sources, leading to potential data uncertainties.

Strongest · Setup experience 12/20 Weakest · Reliability 5/20

Reliability
5/20
Security and permissions
9/20
Maintenance
6/20
Documentation
10/20
Setup experience
12/20
Why each score
Reliability 5/20
The manifest is well-formed against the 2025-10-17 schema and names a PyPI package with stdio transport; the README declares 9 tools with typed parameters and defaults. However, this is a static review with no source code, tests, or CI workflows supplied, so nothing verifies that the MCP init/tool-list handshake succeeds or that declared tools match real behavior. No error-handling or edge-case evidence exists. Per calibration, reliability cannot exceed 12 without execution evidence, and uncertainty lowers it further; hence a low score of 5.
Security and permissions 9/20
No red-line issues found: no credential handling, no real tokens or secrets in install examples, no observable malware or covert exfiltration code, and no destructive default operations. The main risk is outbound network access to third-party data sources (eastmoney, xueqiu, sina), which is disclosed in the tool parameters. However, there is no security policy, data-flow disclosure, least-privilege discussion, or adversarial-testing evidence, so full marks are not justified. Given the low inherent risk, the score is 9.
Maintenance 6/20
The repository is not archived and carries an MIT license; version 0.3.7 and the PyPI package name imply iterative releases. Yet no commit history, release dates, issue-response activity, dependency-update evidence, or security-response channel was provided. The 215 stars are treated as a discovery signal only and add no points. Missing maintenance evidence deducts, yielding 6.
Documentation 10/20
The README is strong: bilingual, with per-tool parameter tables, defaults, and data sources, an install section covering Smithery, uvx, and local development, plus a complete technical-indicator reference. Gaps include no troubleshooting guide, no rate/usage-limit disclosure, no source-reliability notes, and no known-limitations section. The docs are usable but carry hidden assumptions (e.g., uv installed, third-party endpoints available), matching the anchor at 10.
Setup experience 12/20
Setup is clearly presented with copy-paste commands and ready-made mcpServers JSON for both uvx and local uv invocation, plus a Smithery one-liner; stdio and streamable-http modes and their flags are documented. The steps are few and client-centric. However, no CI workflow or committed tests prove the install works end-to-end, so the static cap of 15 applies; because PyPI publication and tool defaults could not be independently verified, the score is conservatively set at 12.

Static review · not runListed 2026-08-07

Read the FMRS scoring method →

Fit and risk

What it can accessUses the network

Best for

  • Investors and researchers needing China A-share market data
  • Quantitative trading strategy developers and researchers
  • Developers requiring Chinese financial data

Not for

  • Users needing global stock market data
  • Users requiring high-frequency or ultra-low-latency real-time data
  • Users with strict accuracy requirements who lack validation channels (note limitations of data sources)

Required permissions

  • Network access: to fetch data from sources like Eastmoney, Sina, and Xueqiu
  • Local file access: when installed locally, may read/write configuration files and temporary files

Risks and side effects

  • Data delays or temporary unavailability of data sources may cause tool failures
  • Data accuracy may be limited by sources; cross-validation recommended
  • Third-party data sources might have usage terms; ensure compliance

Setup

Before you start

Runtime：Node.js · Python

  1. Install uv if not already installed.
  2. Add the following configuration to your MCP client settings:

{
"mcpServers": {
"akshare-one-mcp": {
"command": "uvx",
"args": ["akshare-one-mcp"]
}
}
}

  1. Alternatively, install automatically via Smithery:

npx -y @smithery/cli install @zwldarren/akshare-one-mcp --client claude

claude_desktop_config.json 
{
  "mcpServers": {
    "akshare-one-mcp": {
      "command": "uvx",
      "args": ["akshare-one-mcp"]
    }
  }
}

Shown for Claude Desktop. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.

.vscode/mcp.json
{
  "servers": {
    "akshare-one-mcp": {
      "command": "uvx",
      "args": [
        "akshare-one-mcp"
      ]
    }
  }
}

Goes in your project's .vscode/mcp.json (VS Code uses a "servers" key).

Terminal
claude mcp add akshare-one-mcp -- uvx akshare-one-mcp

Run it in a terminal; replace any <…> placeholders with your own values first.

Generate and validate client config

Check that it works

Check that tools like get_hist_data and get_realtime_data appear in your MCP client's tool list, then ask 'What is the real-time price of 000001?' — a returned data point confirms the connection works.

Troubleshooting

  1. Check network connectivity and ensure data sources are reachable if tool calls fail
  2. If starting with uv fails, confirm uv is installed and in PATH
  3. For historical data, verify stock code format (e.g., 000001)
  4. When using streamable-http mode, ensure endpoint is http://<host>:<port>/mcp

Things to try

Once connected, you can ask your AI assistant things like:

  • Show me the last 100 daily candlesticks for stock 000001 with forward adjustment
  • Get real-time market data for stock 000001
  • List the 10 most recent news items for stock 000001
  • Fetch the latest balance sheet and key financial metrics for 000001

Tools 9

get_hist_data read-only
Get historical stock market data with support for multiple time periods and adjustment methods.
get_realtime_data read-only
Get real-time stock market data.
get_news_data read-only
Get stock-related news data.
get_balance_sheet read-only
Get company balance sheet data.
get_income_statement read-only
Get company income statement data.
get_cash_flow read-only
Get company cash flow statement data.
get_inner_trade_data read-only
Get company insider trading data.
get_financial_metrics read-only
Get key financial metrics from the three major financial statements.
Show 1 more tools
get_time_info read-only
Get current time with ISO format, timestamp, and the last trading day.

Use cases

Retrieve historical stock data for analysis and backtesting
Get real-time data for monitoring or algorithmic trading
Access financial statements for fundamental analysis
Fetch stock-related news for sentiment monitoring

Supported clients

Claude Desktop

Listed from the project's documentation, not tested by this site.

Overview

akshare-one-mcp is an MCP server based on akshare-one, providing comprehensive interfaces for China stock market data. It offers a set of powerful tools for retrieving financial information including historical stock data, real-time data, news data, and financial statements.

Similar servers

Telegram MCP Server 70 · B

Telegram MCP server powered by Telethon to let MCP clients read chats, manage groups, and send/modify messages, media, contacts, and settings.

★ 1.8k · Tools 76 Compare with this →
Compare with other MCP servers

Source revision 27678a7239d3 Data synced 2026-10-12 Read the FMRS scoring method