---
title: "Integrations — FireFight"
description: "Connect 53 tools across cloud and hosting, observability, on-call, databases, code and issue tracking, or any MCP server, and decide what each one may do."
canonical: https://firefight.app/product/integrations
---

# Integrations

**Your whole stack, behind one set of rules**

Connect your cloud, observability, databases, code and on-call tools. Halon and your agents use them to work through an incident, and you decide what they may change.

[Sign in with Slack](https://app.firefight.app/)

## Connected once, governed every time

Connect a tool once. From then on, every call to it follows your rules.

### Connect

**Connect in seconds.** Sign in to many tools with one click. Any tool with an MCP server connects too.

- One click sign in
- Any MCP server
- Encrypted in transit and at rest

### Choose

**See everything a tool can do, and switch off anything you never want touched.** You decide who can use each one.

- Every capability listed
- Switch off any one
- Turn a connection off at once

### Govern

**You decide who can see and who can change each tool, down to production.** Agents only get the access you give them, and every call is on the record.

- Access per person and per environment
- Approval where you want it
- Every call recorded

## 53 integrations across your whole stack

Not listed? Connect any MCP server and choose what it may do.

- **Cloud and hosting** (Where your apps run, and what you can change there): AWS, Azure, Cloudflare, Convex, DigitalOcean, Fly.io, Google Cloud, Kubernetes, Modal, Netlify, Northflank, Railway, Render, Trigger.dev, Vercel
- **Observability** (Logs, metrics, traces and errors for investigations): Axiom, Better Stack, Datadog, Grafana, Honeybadger, Honeycomb, Logfire, New Relic, OpenStatus, Sentry, SigNoz
- **Alerting and on-call** (Who is on call, and the alerts and incidents paging them): Opsgenie, PagerDuty
- **Status pages** (What your customers see about an incident or planned work): Instatus, Statuspage
- **Product analytics** (Who an incident reaches, what they saw, and the feature flags behind it): Mixpanel, PostHog
- **Databases** (Your data and how it is set up): ClickHouse, Neon, PlanetScale, PostgreSQL, Supabase, Tinybird, Turso, Upstash
- **Code** (Correlate incidents with what shipped): Azure DevOps, Bitbucket, GitHub, GitLab
- **CI** (Builds, tests and their logs, and rerunning them): CircleCI
- **Coding agents** (Agents that write a fix's code change and open the pull request): Cursor, Devin, Factory
- **Issues** (Tickets and follow-up work tied to incidents): Jira, Linear
- **Knowledge** (Runbooks and documentation as context): Confluence, Notion
- **Custom** (Anything with an MCP server): Custom MCP server

## Every tool, one switch away

Every connected tool in one place. Turn any connection off in one click.

## Built for real production setups

Separate environments, multiple accounts, and a clear view of what runs where.

### Separate access per environment

Give production and development their own credentials, so each gets only the access it needs.

### Every account, side by side

Connect two GitHub organizations or two cloud accounts, each under its own name.

### Your infrastructure, mapped

What your clouds, hosts, databases and code hosts run appears on the resource map, each item linked to its page at the provider.

### Metrics as charts

When Halon reads metrics, you see the charts under its answer, in the dashboard and in Slack.

### Code Halon never touches

Protect paths such as .github/workflows/ on each GitHub, GitLab, Bitbucket and Azure DevOps connection. Halon's code changes stay out of them.

### Know when a connection breaks

Each connection shows whether it is healthy, with the error when it is not.

## Documentation

- [Connect an integration](https://firefight.app/docs/api/integrations.md): Connecting, choosing capabilities, environments and several accounts.
- [What Halon can do](https://firefight.app/docs/ai/tools-and-approvals.md): What Halon does with each group of tools, and when it asks first.
- [Permissions](https://firefight.app/docs/gateway/permissions.md): Who can use each tool, and how to change it.
- [Approvals](https://firefight.app/docs/gateway/approvals.md): Hold a change until a person approves it.
- [Status pages](https://firefight.app/docs/api/status-pages.md): Connect Statuspage or Instatus and keep customers told.
- [Issue tracking](https://firefight.app/docs/workspace/issue-tracking.md): Open follow-up work in Linear or Jira from an incident.

## Connect the tools you already run

Sign in with Slack, then connect your first tool in a click. [Sign in with Slack](https://app.firefight.app/).