Introducing Agent Native
Observability
Connect SigNoz to your coding agents (e.g. Claude Code, Cursor) and debug production issues without leaving your dev environment.
Traces, logs, metrics, service topology, and your actual codebase - all in one place. Or use Noz, our new AI teammate out-of-the-box.
Powerful, In-context Observability
In the tools you need. At the time you need.
SigNoz MCP Server
Plug into Claude Code or Cursor in minutes. Get full observability context in every session: traces, logs, metrics, service topology, and deployment history. Start debugging in your terminal.
Noz : SigNoz AI Teammate
Your AI teammate inside SigNoz. Ask about your logs, traces, and metrics in plain English, and Noz investigates across your data, explains what it finds, and suggests what to do next. It can also create dashboards, alerts, and views for you.
Why Agent Native Observability
Talk to your observability stack in natural language.
Describe what you want in natural language and SigNoz builds it. Create dashboards from 80+ templates or from scratch. Generate alerts with sensible defaults. Run queries and get results in chat, no query language required.
“Create a latency dashboard for my payment service” → Done. Fully interactive.
Learn more
From alert to root cause to fix. In one session.
Connect SigNoz to your coding agent and get the full picture in one query: traces, logs, metrics, service topology, and deployment history. Your codebase is already loaded, so you can correlate production telemetry with the code that caused it.
Connect to Kubernetes, Git, and Jira through your cloud provider CLI in your dev environment. Create issues, diagnose root cause, and ship a fix without leaving your env.
Learn moreTeach it your team's debugging heuristics. Not someone else's.
Install the SigNoz plugin in your coding agent. Claude Code, Cursor, Codex, or Gemini can then work in SigNoz from your editor: look up docs, query your data, build dashboards, and create and triage alerts.
Then add your team's judgment. Put your runbooks, service topology, and escalation paths in a skills.md, share it in a GitHub repo, and the agent debugs the way your team does.
Learn more
Plugged into your tribal knowledge. No third-party detours.
Connect to the tools your team already uses, like GitHub, Jira, and Kubernetes, with no routing through third-party connectors or external networks. Agents use your existing runbooks, docs, and best practices to pinpoint issues fast. Team knowledge stays in the systems your team already trusts.
Most AI SRE tools hit an adoption wall here: limited access to internal tools means limited reasoning. With SigNoz MCP and coding agents, you work inside your environment, not around it.
Watch VideoOpen by design. No lock-in. Ever.
SigNoz is a neutral data layer for AI-assisted debugging, built on open standards. It uses OpenTelemetry - the CNCF standard, so your instrumentation, your agents, and your investigation data stay yours. Extend across vendors, port across tools, and define your own debugging workflows to share as reusable skills. The open investigation format SigNoz uses becomes a standard your team can build on.
