KLinePic MCP Server
Generate annotated post-trade candlestick review-chart PNGs from broker or exchange fills. Includes capability checks, no-charge preflight validation, and chart rendering; no signals or order execution.
README
KLinePic MCP Server and Agent API Examples
Turn broker or exchange fills into annotated candlestick review-chart PNGs. AI agents can discover capabilities, validate a request without spending chart quota, and render the finished chart through three MCP tools.
KLinePic is for post-trade review and visualization. It does not provide trading signals, price predictions, investment advice, or order execution.
What is included
| Tool | What it does | Quota impact |
|---|---|---|
klinepic_get_capabilities |
Checks API-key permissions, plan limits, remaining quota, and supported endpoints. | None |
klinepic_preflight_review_chart |
Validates fills, K-line data, scopes, and estimated cost before rendering. | None |
klinepic_create_review_chart |
Returns an annotated candlestick PNG as native MCP image content. | Uses one chart render |
The repository also includes:
- A working JavaScript stdio MCP server.
- A small Agent API client used by the MCP tools.
- Broker, Binance, US-stock, futures, and MetaTrader sample data.
- A complete review-chart request and a no-charge preflight example.
- Docker support, an OpenAPI mirror, and machine-readable discovery files.
Verified functionality
Local installation was last verified on 2026-07-14. Live API capabilities and preflight were last verified on 2026-07-13:
| Check | Result | What was proved |
|---|---|---|
| API adapter tests | PASS — 4/4 | Bearer authentication, forced review-chart mode, PNG-to-base64 conversion, and structured error handling. |
| Real MCP stdio handshake | PASS — 3 tools | A real MCP client completed initialization and discovered all three tool schemas. |
| Fresh Cline-compatible install | PASS — 3 tools | The Windows cmd /c npx configuration launched the GitHub package from an empty temporary directory and completed MCP initialization plus tool discovery. |
| One-click MCPB bundle | PASS — 3 tools | The official MCPB 0.3 validator accepted the manifest; the packed bundle included runtime dependencies and completed a real stdio handshake. |
| Live production capabilities | PASS | The MCP server called the production capability endpoint with a scoped Agent API key. |
| Live production preflight | PASS | The sample request passed production validation without rendering a chart or consuming chart quota. |
| Container build | PASS in CI | GitHub Actions builds the Docker image from a clean checkout. |
Reproduce the local checks:
npm ci
npm test
npm run test:mcpb
Run the live, no-charge check with your own key:
KLINE_AGENT_API_KEY=kline_agent_xxx npm run test:production-preflight
The live check calls capabilities and preflight only. It does not call the paid render tool.
Quick start
Requirements:
- Node.js 20 or newer.
- A KLinePic Agent API key. Create one from the account page; the free tier does not require a sales conversation.
One-click install for MCPB-compatible desktop clients:
Download klinepic-mcp-server-0.1.0.mcpb, open it in the desktop client, and enter the Agent API key when prompted. The key field is marked sensitive. The API origin defaults to https://klinepic.com; change it only to a trusted HTTPS endpoint because the key and full trade payload are sent there.
Run directly from GitHub:
KLINE_AGENT_API_KEY=kline_agent_xxx \
npx -y github:sher1096/klinepic-agent-api-examples#c48f3dc12a21834a2af165c223230f0cf9500fef
Or clone the repository:
git clone https://github.com/sher1096/klinepic-agent-api-examples.git
cd klinepic-agent-api-examples
npm ci
KLINE_AGENT_API_KEY=kline_agent_xxx npm start
MCP client configuration
Most macOS and Linux clients can run the GitHub package directly:
{
"mcpServers": {
"klinepic": {
"command": "npx",
"args": [
"-y",
"github:sher1096/klinepic-agent-api-examples#c48f3dc12a21834a2af165c223230f0cf9500fef"
],
"env": {
"KLINE_AGENT_API_KEY": "kline_agent_xxx"
}
}
}
}
Some Windows clients need cmd.exe to resolve npx.cmd:
{
"mcpServers": {
"klinepic": {
"command": "cmd",
"args": [
"/c",
"npx",
"-y",
"github:sher1096/klinepic-agent-api-examples#c48f3dc12a21834a2af165c223230f0cf9500fef"
],
"env": {
"KLINE_AGENT_API_KEY": "kline_agent_xxx"
}
}
}
}
Keep the key in the client environment or secret store. Do not commit it.
Docker
Run the published image:
docker run --rm -i \
-e KLINE_AGENT_API_KEY=kline_agent_xxx \
ghcr.io/sher1096/klinepic-mcp-server:0.1.0
Or build it locally:
docker build -t klinepic-mcp .
docker run --rm -i \
-e KLINE_AGENT_API_KEY=kline_agent_xxx \
klinepic-mcp
The container can start without a key so a registry can inspect its tool schemas. API calls still require a valid key.
Recommended agent flow
- Call
klinepic_get_capabilities. - Build a request from the user's fills and K-line data.
- Call
klinepic_preflight_review_chart. - Fix any validation or permission error.
- Ask for confirmation before a quota-consuming render.
- Call
klinepic_create_review_chartand return the PNG.
Preflight first: it catches malformed CSV, unsupported sources, missing scopes, IP restrictions, plan state, and quota limits without rendering.
Sample data
| File | Use case |
|---|---|
review-chart-request.json |
Complete request for preflight or rendering |
broker-trades-template.csv |
Generic broker exports |
binance-trade-history-template.csv |
Binance trade history |
us-stock-trading-journal-template.csv |
US-stock journals |
futures-trade-review-template.csv |
Futures reviews |
metatrader-trade-history-template.csv |
MetaTrader 4/5 history |
Run the standalone no-charge example:
KLINE_AGENT_API_KEY=kline_agent_xxx node examples/preflight.mjs
Troubleshooting
KLINE_AGENT_API_KEY is required— create a key and pass it through the MCP client's environment.- HTTP 401 — the key is invalid or revoked.
- HTTP 402 — the account or key quota is exhausted; capabilities shows the relevant limit.
- HTTP 403 — the key lacks the required scope or its IP allowlist blocks the caller.
- HTTP 422 — the request data is invalid; run preflight and inspect the structured details.
- Windows cannot find
npx— use thecmd /c npxconfiguration above.
Documentation and discovery
- Cline installation guide
- Standalone field guide
- Complete AI/GEO context bundle
- Human-readable Agent API guide
- OpenAPI contract
- GitHub Pages OpenAPI mirror
- Copy-paste quickstart
- Accepted one-click MCPB install and checksum Q&A
- Accepted official MCP Registry and OCI Q&A
- Claimed Glama listing and quality score
- MCP Repository directory listing
- Accepted MCP installation Q&A
- Accepted broker-fill workflow Q&A
- 中文 A 股交割单问答
Security and scope
- Report suspected vulnerabilities through the private security-reporting form, not a public issue.
- API keys are sent only to
https://klinepic.comunlessKLINEPIC_BASE_URLis explicitly overridden. - The server does not store keys, trade records, or generated images locally.
- Use a dedicated key with the smallest useful scopes and limits.
- Review data-handling requirements before sending real trading records.
License
The MCP adapter and examples in this repository are available under the MIT License. The hosted KLinePic API is a separate service governed by its published terms and privacy policy.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.