# SigNoz MCP

> SigNoz MCP lets you manage observability alerts and rules directly through your AI agent. This open-source alternative handles infrastructure monitoring, allowing you to list, create, and update alert thresholds without touching a complex web console.

## Overview
- **Category:** cloud-infrastructure
- **Price:** Free
- **Tags:** alerting, infrastructure-monitoring, open-source, telemetry, incident-response

## Description

Connecting SigNoz to your AI agent gives you full control over your monitoring stack using natural language. If you rely on observability data but hate navigating dozens of dashboards just to check an alert threshold, this is for you. You can tell the agent exactly what rules exist, update a condition on a high-priority metric, or delete old alerts that are no longer relevant. It's about treating your entire monitoring infrastructure like a conversation with your team member—fast, direct, and actionable. When you connect SigNoz via Vinkius, you gain instant access to this powerful set of tools alongside thousands of others, letting you keep all your operational data centralized right from your code editor.

## Tools

### create_rule
Creates a brand new alert rule using specified details like the metric, condition, and thresholds.

### delete_rule
Removes an entire alert rule from the monitoring system by its identifier.

### get_service_account_me
Confirms your current service account credentials and associated API key permissions.

### list_rules
Retrieves a comprehensive list of all currently configured alert rules in the system.

### update_rule
Modifies an existing alert rule, allowing you to change its thresholds or triggering conditions.

## Prompt Examples

**Prompt:** 
```
List all alert rules in my SigNoz instance.
```

**Response:** 
```
I've retrieved 8 alert rules. Some active rules include 'High CPU Usage' (ID: 1), 'Database Latency' (ID: 2), and 'Service 5xx Errors' (ID: 3). Would you like to see the details for any specific rule?
```

**Prompt:** 
```
Show me the details of my current service account.
```

**Response:** 
```
Your API key is valid. You are connected as 'Monitoring-Admin' with the following permissions: [Alerts: Read/Write, Dashboards: Read]. Your account ID is 'sa-99283'.
```

**Prompt:** 
```
Delete the alert rule with ID 'alert-123'.
```

**Response:** 
```
The alert rule 'alert-123' has been successfully deleted from your SigNoz instance.
```

## Capabilities

### Check account status
Verify your current service account details and API key permissions instantly.

### List existing rules
Retrieve a comprehensive list of all alert configurations currently in use.

### Create new alerts
Build entirely new alert rules, setting specific conditions and thresholds for monitoring.

### Modify existing rules
Update the thresholds or trigger conditions of any rule that's already running.

### Remove unused alerts
Quickly delete alert rules to keep your monitoring environment clean and focused.

## Use Cases

### Debugging an unexpected outage
A developer notices high latency spikes but isn't sure which monitoring rule is failing. They ask their agent, 'List all rules related to database performance.' The agent uses `list_rules` and points them directly to the relevant alert ID, saving a deep dive into multiple dashboards.

### Onboarding a new microservice
A platform team needs to monitor a freshly deployed service. They use their agent to execute `create_rule`, defining the specific metrics and thresholds needed for the new component, ensuring zero monitoring gaps from day one.

### Compliance cleanup
An SRE is tasked with auditing old alerts that haven't been touched in six months. The agent uses `list_rules`, allows the engineer to review the list, and then executes `delete_rule` for every flagged obsolete alert.

### Validating credentials before a release
Before pushing code that relies on monitoring data, an engineer first calls `get_service_account_me`. The agent confirms the API key is valid and shows current Read/Write permissions, preventing deployment failures later.

## Benefits

- Audit your entire stack instantly. Instead of clicking through complex UIs to find all alert configurations, you simply tell the agent to list all configured rules.
- Automate rule changes during deployment. Use the `create_rule` tool within your CI/CD process to ensure new services have monitoring coverage immediately.
- Keep things clean. Run a check for stale or redundant alerts and use `delete_rule` to remove old configurations, keeping your dashboard noise low.
- Validate connectivity instantly. Before writing complex code, run `get_service_account_me` to confirm your API key is valid and you have the correct permissions.
- Adjust thresholds on the fly. If a service's normal operating range shifts, use `update_rule` to change its specific alert conditions without manual console access.

## How It Works

The bottom line is that you manage your entire monitoring stack using plain conversation, not clicking through complex web UIs.

1. Subscribe to this MCP in Vinkius, then enter your SigNoz Instance URL and API Key.
2. Call the agent and tell it what you need; for example, 'list all alert rules.'
3. The agent executes the command, retrieves the data, and presents the actionable results back to you.

## Frequently Asked Questions

**How does the SigNoz MCP handle different alert metrics?**
The MCP uses specific JSON payloads to define which metric, condition, and thresholds are needed for a new rule. You don't need to worry about the payload structure; just tell your agent what you want monitored.

**Can I use SigNoz MCP to delete an alert rule?**
Yes. If you know which rule needs removing, you can use `delete_rule` through your agent. It handles the necessary cleanup in SigNoz for you.

**Do I need special permissions to run list_rules?**
The MCP first runs `get_service_account_me` to validate your connection. If your service account lacks read access, the agent won't be able to retrieve the rule list.

**Is SigNoz MCP only for new alerts?**
Not at all. You can use `update_rule` to modify any existing alert—you don't have to recreate it just because you want a slightly different threshold.

**What happens if I try to create a rule with bad data?**
The agent will catch the error before sending it. It reads the tool documentation and tells you exactly what format the metric, condition, or thresholds must be in, so you can correct your request.