Skip to main content
Promptster runs alongside Claude Code and Cursor during a coding assessment, capturing a detailed telemetry record of the session: prompts, file changes, commands, test runs, and architecture decisions. Hiring teams get an objective evidence trail to evaluate engineering judgment — not just final output.

Candidate Quickstart

Install the CLI, redeem your key, and start your assessment in under 5 minutes.

Hiring Team Quickstart

Create your first assessment and generate candidate keys in minutes.

CLI Reference

All Promptster CLI commands for candidates — start, submit, explain, and more.

API Reference

REST API for managing assessments, sessions, and reviewing candidate data.

​
How it works

1

Hiring team creates an assessment

Define the task brief, role, and time limit. Generate candidate keys to distribute.
2

Candidate redeems their key

Run promptster start PST-XXXX-XXXX. The CLI configures Claude Code or Cursor to capture session telemetry automatically.
3

Candidate completes the task

Work normally using AI coding tools. Promptster captures prompts, file diffs, commands, and architecture decisions in the background.
4

Candidate submits

Run promptster done to flush all events and close the session.
5

Hiring team reviews

Browse the session timeline, review captured decisions, and examine derived signals to evaluate engineering judgment.
Promptster only captures what happens inside your AI coding tool session. Keystrokes, clipboard contents, browser activity, and environment variables are never recorded. See Privacy & Data for the full list.