naver-stock-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| MCP_HOST | No | HTTP bind host. | 127.0.0.1 |
| MCP_PATH | No | HTTP endpoint path. | /mcp |
| MCP_PORT | No | HTTP bind port. | 3000 |
| NO_PROXY | No | Hosts that bypass the proxy. | |
| HTTP_PROXY | No | Proxy for outbound HTTP requests to Naver. Lowercase names and user:pass@ credentials are supported. | |
| DOTENV_PATH | No | Path of the env file to load. | .env |
| HTTPS_PROXY | No | Proxy for outbound HTTPS requests to Naver. Usually the only proxy variable needed since all Naver endpoints use HTTPS. Lowercase names and user:pass@ credentials are supported. | |
| MCP_TRANSPORT | No | Transport mode: stdio or http. | stdio |
| NAVER_TIMEOUT_MS | No | Request timeout in milliseconds. | 15000 |
| NAVER_USER_AGENT | No | User-Agent override for requests to Naver. | Chrome UA |
| MCP_ALLOWED_HOSTS | No | Allowed Host headers, comma-separated. Required when binding to 0.0.0.0 to keep DNS rebinding protection. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| search_researchA | Search brokerage research reports on Naver Stock (stock.naver.com/research). Reports are mostly in Korean. Every result has detailUrl (the Naver detail page); set includePdf=true to also get pdfUrl (the original PDF). Results are sorted newest first. Categories: market = market commentary; invest = investment strategy; company = company/stock analysis; industry = industry analysis; economy = economic analysis; debenture = bond/fixed income. For reports on specific stocks use category=company with stocks; for specific sectors use category=industry with industries. query is a keyword search over titles and summaries (Korean keywords work best). |
| get_researchA | Get one research report: full summary text, original PDF URL (pdfUrl) and Naver detail page (detailUrl). Company reports also include the target price, previous target price and the stock price on the publication date. Take category and nid from search_research results. |
| find_stockA | Look up stock codes by name using Naver Stock autocomplete. Korean stock names work best. |
| list_brokersA | List the brokerages and institutions that publish research, with their codes. |
| list_industriesA | List the industry keys and Korean names usable as filters for category=industry. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 5 tools
Each tool has a clearly distinct role: search_research vs get_research separate discovery from detail retrieval, and find_stock, list_brokers, and list_industries are non-overlapping lookup helpers. An agent can easily select the right tool for each task.
All names follow a consistent snake_case verb_noun pattern (search_research, get_research, find_stock, list_brokers, list_industries). The verbs are standard and predictable.
Five tools are well-scoped for a research-report browsing server: two core operations plus three supporting lookups. No tool feels redundant or extraneous.
The research workflow is covered end-to-end: search by category/keyword, fetch full report details with PDF, and resolve stock/broker/industry filters. Minor gap: no tool for current stock quotes or price history, though the server is focused on research reports.