- Home
- Skills
- Trading & Investing
- Stock MCP: A-Share, HK & US Stock Research
Works with the AI Agents you already use
Stock MCP: A-Share, HK & US Stock Research
A-share, HK & US stock research: data since 1990, A-share 60-minute bars, SQL screening, Python backtests and email reports. Separate MCP key required.
Free
Stock MCP: A-Share, HK & US Stock Research
Example session with this skill installed
Help me set up Stock MCP for A-share stock research and 60-minute candlestick analysis. Explain how I can later backtest and email a report. Give me a configuration template without using a real key or running market queries.
- Read your context and instructions
- Compiled the stock mcp
Stock MCP setup
Connect your agent to the hosted financial data service using these settings:
| Setting | Value |
|---|---|
| Server name | stock-mcp |
| Transport | Streamable HTTP |
| Endpoint | https://api.zhiweigo.com/mcp |
| Authentication | Authorization: Bearer YOUR_MCP_KEY |
| Account and key management | https://zhiweigo.com/ |
- Sign in at zhiweigo.com and open Personal Center > MCP > Create Key.
- Select the tool permissions needed for your workflow.
- Store the key privately in your agent's MCP connection settings, reconnect, and retrieve the available tool list.
For clients that accept this configuration format:
{
"mcpServers": {
"stock-mcp": {
"type": "streamable_http",
"url": "https://api.zhiweigo.com/mcp",
"headers": {
"Authorization": "Bearer YOUR_MCP_KEY"
}
}
}
}
Client schemas differ; use your client’s documented remote HTTP settings if this format is not supported.
Choose the tools for your workflow
| Task | Tools |
|---|---|
| Discover schemas and screen stocks | stock_data |
| Analyze A-share 60-minute bars | stock_data; sandbox_exec for larger analysis |
| Run Python backtests and produce charts | sandbox_exec, sandbox_keepalive, sandbox_files |
| Add public news and sources | web_search, web_reader |
| Email a requested research summary | send_email |
| Check time and available credits | current_time, credit_balance |
Historical coverage spans 1990 to the present, but each market, security and bar frequency needs its own coverage check. Do not assume the 60-minute data starts in 1990. Before a backtest, verify the price adjustment basis, available announcement dates, fees and slippage. Email delivery uses your account's linked address and is performed only when you request it.
Next step
After connecting, try: “Confirm the available 60-minute data for my chosen A-share stock over the last month. Report the actual date range, price/volume summary and any missing intervals.”
This is a setup walkthrough: no market query, backtest or email has been executed. The skill is free; data calls and cloud computation follow the provider's credit pricing.
Connects securely to your tools. The creator never sees your data.
About this skill
Stock research from question to report
Stock MCP (股票MCP / 股票 MCP) helps your AI agent research China A-shares, Hong Kong and US stocks, plus indices, using the hosted financial data service at zhiweigo.com.
Historical data spans 1990 to the present, including China A-share 60-minute candlestick data. Coverage and freshness vary by market, security, metric and frequency. The skill checks the actual available range before analysis.
What you can do
- Screen stocks with SQL: turn investment criteria into read-only queries, compare financial statements and rank companies with clear data dates.
- Study prices and fundamentals: examine daily prices, A-share 60-minute bars, valuations, profitability, cash flow and index benchmarks.
- Backtest in cloud Python: evaluate strategies in an isolated sandbox and generate equity curves, drawdowns, comparison charts and report files.
- Add public news: combine structured financial data with relevant web sources and citations.
- Send research emails: on your request, send an HTML research summary to the email address linked to your Stock MCP account.
What makes the workflow useful
The skill tells the agent to inspect schemas instead of guessing columns, distinguish reporting dates from publication dates, explain unadjusted prices, include trading costs in backtests and disclose missing data. It helps turn a broad stock question into an analysis you can inspect and reproduce.
What is included
A portable SKILL.md workflow, a connection guide, and a reference with bounded SQL discovery queries and financial table navigation. No executable installer or embedded API key is included.
Setup and pricing
The skill is free. The hosted data and compute service is separate and uses credits. Create your own key at zhiweigo.com, connect your agent to https://api.zhiweigo.com/mcp using Streamable HTTP, and configure the Authorization header privately. Tool availability depends on your key permissions. Installing the skill does not automatically connect the MCP server.
Use an agent that supports SKILL.md and remote MCP with custom authentication headers, directly or through a supported bridge. Configuration varies by client and version.
Try it
“Find A-share companies profitable in each of the last three complete fiscal years, rank them by annual ROE, and show the criteria and data dates.”
“Inspect the available 60-minute bars for an A-share stock, analyze price and volume, and explain missing intervals.”
“Backtest a moving-average strategy against buy-and-hold, state costs and adjustment assumptions, then email the research summary to my linked account address.”
Research outputs are for analysis and decision support. The skill does not place trades or guarantee returns.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Trading & Investing.
Requires a separately configured Stock MCP connection and usage credits. Tools depend on key permissions. Data coverage and freshness vary; 1990-to-present coverage does not apply to every security or intraday series. Daily prices are unadjusted. Backtests depend on available historical inputs. Email goes only to the current account linked address. No trade execution or guaranteed returns.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 7 days ago
- Free with an Agensi account
- Accounts
Zhiweigozhiweigo.com accountStock MCP API key
- Software
- Remote MCP client
- Optional
- Stock MCP sandbox accessStock MCP email access
Free skill; hosted MCP calls use credits under zhiweigo.com pricing. Set up your own MCP connection privately. Installing the skill does not include API access or credits.
Permissions
Allowed Hosts
File Scopes
Network calls go to api.zhiweigo.com; optional search tools retrieve public web sources. Python commands and generated files use the provider cloud sandbox (/workspace). Email is only user-requested, to the linked account address. No local shell access is required for normal use.