Quick Start

​

Set Up Annie

1

Create Your Account

Sign up and create your Anyshift workspace, or reuse your existing project.

Create Account

Get started for free

2

Connect One Integration

Reuse your existing integrations, or connect a first data source. An incident investigation needs the relevant infrastructure and health signals; one connected source may not cover the whole task.

GitHub

Connect a repository

AWS

Connect an account

Kubernetes

Install the agent in your cluster with Helm

Datadog

Connect monitoring

For a quick test, connect a dev account first. You can add production later.

3

Talk to Annie

Ask Annie a question about your infrastructure:

Web Chat

Chat directly in the dashboard

Terminal (CLI)

Ask from your terminal

Slack

Mention @Annie in Slack

IDE (Annie Remote MCP)

Use in your AI coding assistant

To give your own coding agent production context, use the plugin or Graph MCP. The plugin already includes Graph MCP. For an example, review a Kubernetes permission change.

4

Set Up Incident Monitoring (Optional)

Connect an alerting tool and define which events should start an investigation. Annie compares the alert with available evidence and reports findings or missing data. Choose the result destination; Slack delivery requires a configured Slack connection and channel.

PagerDuty

Investigate matching incident events

Datadog

Auto-RCA on monitor alerts

Slack on-call

Register alert bots for on-call monitoring

Incident.io

Connect incident management

To enable automatic RCA in Slack, invite @Annie to your alert channel and use /register_annie_on_call <bot_slack_id> to register your monitoring bot. See Slack Integration for full setup details.

​

Example First Questions

• “What resources do I have?”
• “Show me the dependencies for [service-name]”
• “What changed in the last 24 hours?”
• “Why did this alert fire?”