Skip to main content

Documentation Index

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

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

What separates a good coding agent run from a great one is the developer experience around it: a fast, reliable feedback loop. Niteshift gives every task a cloud environment with dependencies installed, dev server running, and a browser ready, alongside a live preview to collaborate with the agent as it works.

​
What a task is

A task is a coding session running in its own cloud environment. When you start one, Niteshift clones your repository, runs your Niteshift setup script, boots your dev server, and starts the agent (Claude Code or Codex) inside the environment. The agent edits files, runs commands, and exercises your app the same way you would locally. You watch it happen in the workspace: chat with the agent on the right while the preview, diff, terminal, and logs update on the left. When the work looks good, the agent commits, pushes a branch, and opens a pull request.

​
Where tasks come from

You can start a task from:
  • The web UI at niteshift.dev
  • A GitHub PR or issue comment with /niteshift <prompt>
  • Slack by mentioning @Niteshift
  • A Linear issue assigned to Niteshift
  • An automation triggered by a webhook or schedule
Tasks behave the same regardless of where they start. A Slack-triggered task gets the same environment, the same agent options, and the same workspace as one launched from the web.

​
What the environment gives the agent

Because the dev server is already running and reachable at a preview URL, the agent can verify changes by hitting the same endpoints you would. It can open the preview in a real browser, click through flows, and capture screenshots. It can connect to a branched database so schema changes don’t touch shared data. It can run Docker services your app depends on.

​
Next

Set up your account and run your first task in the Quickstart.