Evaluate your first production incident

Availability: preview

Start by connecting a production monitoring signal to a project and inspecting its incident evidence, diagnosis, and proposed next steps. The current journey is intended for source evaluation.

Prepare the incident application

The current source-evaluation journey runs the incident web application from this repository. It requires PostgreSQL for business data, Redis for login sessions, one bootstrapped login identity, and project configuration. These are incident-application prerequisites, not dependencies of the neutral Harness contracts.

1. Confirm the incident application prerequisites.
2. Read the installation status before choosing an environment.
3. Run migrations and bootstrap the login identity.
4. Create and configure the first project.
5. Send a supported signal and inspect the first incident.

This path is suitable for evaluation and development validation, not a supported customer installation.

For developers: Local Agent Harness

The separate Local path accepts a JSON event, executes it through the shared Harness, and writes inspectable state without PostgreSQL, Redis, or an account. The source preview now documents build, run, inspect, resume, and operator recovery commands. See Run the Local Agent Harness. This development path has different prerequisites and release guarantees from the incident application.

Extend from source

Read Extend the Harness for the implemented tool, policy, Profile, evaluator, artifact, and RunStore contracts. That guide documents source APIs; it is not a claim that a public SDK or generic service UI has shipped.