VersusControl/versus-incident

What is it?

Key features

• 01AI agent learns normal log patterns and raises incidents for new anomalies.
• 02Accepts webhooks from Alertmanager, Grafana, Sentry, CloudWatch SNS and FluentBit.
• 03Sends incidents to Slack, Microsoft Teams, Telegram, Viber, Email and Lark.
• 04Escalates unacknowledged incidents to PagerDuty, Opsgenie, incident.io, ServiceNow or AWS Incident Manager.
• 05Customizable alert messages using Go templates.
• 06YAML configuration with environment variable support.

Use cases

• Consolidating alerts from several monitoring tools into one notification pipeline.
• Detecting unexpected errors in application logs without maintaining alert rules.
• Running on-call escalation for a team that uses PagerDuty or Opsgenie.
• Monitoring Kubernetes clusters and reviewing workloads and Helm releases from one dashboard.

Quick start

1. 1

   Start Redis for the agent's cursor storage

   docker run -d --name versus-redis -p 6379:6379 redis:7
2. 2

   Run Versus in training mode with the AI agent enabled

   docker run -p 3000:3000 -e GATEWAY_SECRET=change-me -e AGENT_ENABLE=true -e AGENT_MODE=training -e REDIS_HOST=host.docker.internal -e REDIS_PORT=6379 -v $(pwd)/config:/app/config -v $(pwd)/data:/app/data ghcr.io/versuscontrol/versus-incident
3. 3

   Open the admin dashboard at localhost:3000 to review learned patterns

Things to know

• !The agent requires a config.yaml and agent_sources.yaml pointing at your logs before it can run.
• !Only the file storage backend is implemented today; redis and database are stubs.
• !The AI agent also requires a running Redis instance for source cursors.

Topics

FAQ

What is versus-incident?

A self-hosted AI SRE agent that learns normal log patterns, auto-raises incidents for new anomalies, and routes alerts to Slack, Teams, Telegram and on-call tools. Versus Incident is a self-hosted incident management service for operations teams. It can create incidents in two ways: an AI SRE agent that reads log sources such as Elasticsearch and local files, learns normal patterns, and opens an incident when a new error appears; or webhook alerts from tools like Alertmanager, Grafana, Sentry, CloudWatch SNS and FluentBit, formatted with Go templates.

Is versus-incident free to use? What license does it have?

versus-incident is open source under the MIT license. Check the license terms before using it in commercial products.

Is versus-incident still maintained?

Actively maintained. The last commit to versus-incident was on Oct 17, 2026.

What can I use versus-incident for?

Consolidating alerts from several monitoring tools into one notification pipeline. Detecting unexpected errors in application logs without maintaining alert rules. Running on-call escalation for a team that uses PagerDuty or Opsgenie. Monitoring Kubernetes clusters and reviewing workloads and Helm releases from one dashboard.

How do I install versus-incident?

Start Redis for the agent's cursor storage: docker run -d --name versus-redis -p 6379:6379 redis:7 Run Versus in training mode with the AI agent enabled: docker run -p 3000:3000 -e GATEWAY_SECRET=change-me -e AGENT_ENABLE=true -e AGENT_MODE=training -e REDIS_HOST=host.docker.internal -e REDIS_PORT=6379 -v $(pwd)/config:/app/config -v $(pwd)/data:/app/data ghcr.io/versuscontrol/versus-incident Open the admin dashboard at localhost:3000 to review learned patterns

Where can I get versus-incident?

The source code and installation instructions are on GitHub: https://github.com/VersusControl/versus-incident