Cloud Dashboard

Documentation Index

Fetch the complete documentation index at: https://docs.selftune.dev/llms.txt

Use this file to discover all available pages before exploring further.

​

Overview

The selftune Cloud dashboard gives you a centralized view of all your connected skills. From here you can review grading signals, manage eval suites, and launch improvement runs.

​

Skill Detail Page

Clicking a skill from the dashboard opens its detail page. This page shows the current snapshot, linked eval suites, and improvement run history.

​

Auto-Setup on First Visit

When you open a skill detail page for the first time — before any eval suites or improve runs exist — selftune automatically:

1. Generates an initial eval suite seeded from the skill’s current behavior
2. Queues your first improvement run using that suite
3. Redirects you to the improve run detail page

This happens without any manual steps. The auto-setup uses your default model configuration. You can adjust model selection (generate, judge, summarize) and apply target before manually queueing subsequent runs.

​

Starting an Improve Run Manually

Once a skill has an eval suite, you can queue improvement runs manually from the skill detail page:

1. Select an eval suite from the dropdown
2. (Optional) Expand Advanced options to choose models and apply target
3. Click Start improve run

You are redirected to the run detail page where you can follow progress in real time.

​

Eval Suite Management

The skill detail page lists all eval suites linked to the source. You can:

• Create a quick suite — auto-generate eval cases from the skill’s current snapshot
• Edit cases — add, remove, or modify individual eval cases before running
• Select a suite — choose which suite to use for the next improve run

See Eval Suites for full configuration options.

​

Improve Run History

Past improve runs appear in the Runs tab. Each entry shows run ID, status, model configuration, and timestamps. Click a run to view detailed output and evolution proposals. See Improve Run Statuses for a breakdown of each status and what it means.