Intelligent operations
Platform capabilities
Security analytics
AI
Dedicated compute
Dedicated compute keeps concurrent queries stable and complex analysis fast.
Learn moreDedicated compute
Dedicated compute keeps concurrent queries stable and complex analysis fast.
Learn moreContact us
Join the community
Try Guance
Start online with usage-based pricing and a true cloud service.
Get startedChoose a Guance edition
Guance MCP Server connects Guance's query, diagnostic, and contextual capabilities to AI IDEs, desktop clients, and automated Agent via Model Context Protocol. Within its authorized scope, AI assistants can call observation capabilities such as logs, metrics, links, events, and infrastructure, while engineers retain the right to review tools, parameters, returned results, and final conclusions.
Guance MCP Server provides standardized tool service interfaces based on Model Context Protocol, supporting AI clients supporting streamableHttp; Actual configuration entry points and fields are subject to each client's documentation.
Trae
Cursor
Windsurf
Claude Desktop
VS Code
CodexAI assistant lacks system on-site:It is difficult to determine the true system state by looking only at the chat context, nor can it verify slow requests, errors, or resource anomalies.
High cost of replicating observation data:Manually pasting logs, traces, metrics, and alerts to AI can easily miss context and pose permissions and privacy risks.
The AI troubleshooting process is not subject to censorship:Without the query tools and parameters, the team finds it difficult to determine whether the conclusions come from real data.
Multi-domain diagnostics require a unified entry point:A single failure often spans logs, metrics, links, events, and infrastructure, requiring AI to query continuously within the same context.
MCP Tool Integration:Opening up Guance capabilities to MCP-enabled AI IDEs, desktop clients, and automated Agent.
Context of the authorization tool:Control AI-available tools, query scope, and data access boundaries through configuration and permissions.
Multi-domain diagnostic capabilities:Assist fault analysis around logs, metrics, links, events, APM, RUM, and infrastructure data.
Human-machine collaborative troubleshooting:AI-generated diagnostic steps and recommendations, engineers review tool calls, data evidence, and conclusions.
MCP Server acts as a protocol bridge between Guance and AI clients, allowing assistants supporting MCP to invoke Guance related capabilities. Teams can incorporate service troubleshooting, alert analysis, trace tracking, and log queries into AI-assisted workflows instead of manually copying data into dialogue boxes.
Once AI tools can call on authorized observation capabilities, they can organize diagnostic steps around specific time windows, service names, Trace IDs, hosts, or alert events. Engineers can still review tools, parameters, returned results, and final conclusions, forming a traceable human-machine collaborative troubleshooting process.
AI Agent can observe AI Agent its own operational processes, while MCP Server allows AI tools to enter Guance data context for assisted analysis. By combining the two, teams can monitor the quality, cost, and risks of AI applications, and also let AI help troubleshoot issues in business systems.
Guance MCP Server is a service entry point that connects Guance query, diagnosis, and contextual capabilities to AI clients supporting Model Context Protocol, helping AI assistants invoke observation capabilities and analyze system issues within authorized scopes.
No. MCP Server is better suited for AI-assisted queries, tool calls, diagnostics, and context organization; The console is still suitable for full viewers, dashboards, alarm configuration, permission management, and team collaboration.
If you mainly work on terminals, you can prioritize using the OWL CLI; If you want AI IDEs, desktop clients, or automated Agent calls to Guance capabilities, you can use MCP Server. Both can also be used together.
Starting from model requests, prompts, response time, errors, and costs, establish observable analytical entry points for large model applications.
Learn about Agent monitoring applications, LLM monitoring applications, session lists, trace lists, and risk event analysis.
Organize logs, metrics, links, events, RUM, and infrastructure queries into executable, auditable diagnostic processes.
Connect AI IDEs, desktop clients, and automated Agent via Model Context Protocol to make recommendations based on authorized observation data.
Want your AI assistant to access your Guance data context?
Contact sales
