1
Create an account and workspace
Go to superlog.sh and click Get started. Sign up with Google, GitHub, or an email and password.Then create a workspace. Your automations, integrations, and members belong to the workspace. You become its admin.
2
Pick a template
Open Automations. Under Start from a template, pick one that matches a task your team does by hand, for example Triage new Sentry issues or Answer support questions.The template fills in the trigger, instructions, and connectors. You can also click Create automation to start blank.
3
Connect the trigger
Each automation needs at least one trigger. If the trigger’s tool is not connected yet, its card shows Connect. Click it, approve access in the pop-up window, and then choose the Slack channels, Sentry projects, or Discord channels to watch.Templates with a Schedule trigger need no connection. Set the frequency and time instead.
4
Add a repository
Under Repositories, click Add GitHub and install the GitHub App on the repositories you want Superlog to use. Then click Add repository and pick at least one. The first repository is the agent’s working directory.
5
Choose a model
In the Agent instructions toolbar, click Choose model and pick a provider and model. Superlog picks a compatible harness for you.Runs use the usage included in your plan unless you add your own API key or ChatGPT subscription in Settings → Models. See Models and harnesses.
6
Save and run
Click Save. The automation is active and runs the next time its trigger fires.To see it work right away, click Run now, or click Test and describe an event for the agent to handle. Open the run to follow the agent’s work live.
Next steps
Connectors
Give the agent access to Slack, Sentry, Datadog, Linear, and your own MCP servers.
Runs
Follow a run, send follow-ups, and post results to Slack.
Tag mode
Mention Superlog in Slack to ask questions about your code and infrastructure.
Invite your team
Add teammates to your workspace.