KnowledgeRadar
Allows searching Bilibili and retrieving content details.
Allows searching GitHub repositories and code via the kr_web_search tool.
Optionally integrates with a local SearXNG instance to provide web search results.
Allows searching Xiaohongshu and retrieving content details.
Allows searching YouTube videos via the kr_web_search tool.
Allows searching Zhihu and retrieving content details.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@KnowledgeRadarSearch for recent academic papers on graph neural networks and give a brief overview."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
KnowledgeRadar
KnowledgeRadar is an MCP server that gives agents a single search and perception layer across general web search, Bilibili, Xiaohongshu, Zhihu, YouTube-via-web, academic metadata, recruitment sites, webpage extraction, research planning, health checks, and runtime task status.
Quick Start
Install Python 3.12 or use the bundled
.python312runtime.Copy
.env.exampleto.envand fill only the providers you want to use.Run
scripts\install.bat.Run
start.cmd.Register
http://127.0.0.1:18765/mcpin your agent, or useconfig\mcp-config-template.json.
Related MCP server: Percival Deep Research
Startup Modes
Mode | Command | Use case |
Foreground HTTP |
| Local debugging; keeps logs visible in the terminal |
Detached HTTP |
| Normal Windows startup from Explorer or scripts |
Hidden HTTP |
| Background startup without a visible console window |
Stdio |
| Agents that launch MCP servers directly |
All startup modes read only the repository-root .env file. It is private local state and is never a source, package, or GitHub input. Logs default to runtime\logs when launched through the included scripts.
User Docs
docs\INSTALL.md: installation and first run.docs\API_KEYS.md: provider keys and validation.docs\MCP_SETUP.md: Agent configuration examples.docs\ACCOUNT_SETUP.md: interactive browser login/profile setup.docs\FAQ.md: common validation states and troubleshooting.docs\SEARXNG.md: optional local SearXNG setup.docs\PRIVACY_AND_LOCAL_STATE.md: what remains local and how public artifacts are checked.
Important Paths
Path | Purpose | Commit |
| MCP server and collectors | Yes |
| Publishable templates | Yes |
| Local account/profile state | No |
| API keys and private runtime settings | No |
| Browser profile links | No |
| Logs, task DBs, decision traces | No |
| Optional bundled Python runtime | No |
MCP Tools
The source of truth is src/server.py @mcp.tool() registrations. Current tools include:
expand_keywordsplan_researchanalyze_decision_logsget_task_statuskr_web_searchsearch_academicextract_web_pageextract_dynamic_pagesearch_bilibilisearch_xiaohongshusearch_zhihusearch_recruitmentget_content_detailget_capabilitieshealth_check
YouTube and GitHub search are exposed through kr_web_search provider modes rather than separate public tools.
Verification
After setup, verify the installed product surface without printing secrets:
.python312\python.exe scripts\verify_api_keys.py
.python312\python.exe scripts\verify_all_capabilities.py --safehealth_check separates main production chains from optional candidates and diagnostic paths. A configured environment should have zero required main-chain FAIL and zero required main-chain NEEDS_INTERACTION in verify_all_capabilities.py --safe; declared optional providers, quota exhaustion, and designed fallback paths may be reported as EXPECTED_DEGRADED without failing the run. Login, QR-code, CAPTCHA, and similar human steps should be reported as NEEDS_INTERACTION until the user completes them.
Secrets
Never commit real API keys, tokens, cookies, browser profiles, SQLite files, or runtime logs. Put private values only in the repository-root .env.
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- Flicense-qualityDmaintenanceProvides comprehensive search capabilities including web search, content extraction, news search, academic search, and AI-powered multi-source research. Enables natural language access to web content and research through a production-ready MCP server.
- Alicense-qualityBmaintenanceAn MCP server that provides autonomous, multi-source web research capabilities for AI agents. It delivers comprehensive, validated information through deep research tools while maintaining security and compatibility with various LLM providers.MIT
- Alicense-qualityFmaintenanceMCP tool server that gives any AI agent the ability to search, scrape, and analyze content across the internet.43MIT
- FlicenseAqualityBmaintenanceA production-grade MCP server that provides a centralized microservice toolkit for LLM agents, enabling web search and extensible tool integration.1
Related MCP Connectors
Hosted MCP with 91 agent tools: X, domains, SEO, Maps, Trends, Search, YouTube, TikTok, and more.
Personal assistant MCP server with search, execute, packages, jobs, secrets, and integrations.
Multi-engine search for AI agents. Trust scoring, local corpus, MCP-native. Self-hostable, BYOK.
Appeared in Searches
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/ga626/knowledge-radar'
If you have feedback or need assistance with the MCP directory API, please join our Discord server