Spectr Documentation

Test everything.
Write nothing.

Spectr is the AI-native testing platform built to help teams scale coverage, eliminate flaky tests, and ship with confidence — without writing a single test by hand.

⚡ Quickstart — 10 min 🔌 SDK Generator 📋 Changelog

REST API

Base URL: https://app.spectrtechnology.com — authenticate with Authorization: Bearer YOUR_API_KEY

POST/api/sdk/runGenerate AI tests for a target URL and step list

GET/api/sessionsList recorded Playwright sessions

POST/api/orchestrateTrigger full LangGraph pipeline (PLAN → GEN → VALIDATE → FIX)

GET/api/reportsFetch test run reports and eval gate scores

POST/api/jira/create-ticketCreate a Jira issue from a test failure

GET/api/sdk/github-actionGenerate GitHub Actions workflow YAML

REST SDKPOST /api/sdk/run

curlpythontypescript

curl -X POST https://app.spectrtechnology.com/api/sdk/run \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "url":       "https://yourapp.com",
    "model":     "claude-sonnet-4-6",
    "testType":  "ui",
    "language":  "playwright-ts",
    "browser":   "chromium",
    "steps": [
      "Navigate to login page",
      "Enter valid credentials and submit",
      "Assert dashboard is visible"
    ]
  }'

Response

{
  "ok":          true,
  "language":    "playwright-ts",
  "plan":        "1. Navigate to login...\n2. Fill credentials...",
  "code":        "import { test, expect } from '@playwright/test';\n...",
  "validation":  "VALID",
  "evalScore":   87,
  "durationMs":  8420,
  "model":       "claude-sonnet-4-6"
}

SDK & CI/CD

Install the TypeScript SDK or use our GitHub Actions workflow generator to wire Spectr into any CI pipeline.

Install

npm install @spectrtechnology/sdk

TypeScriptESM / CJS

import { SpectrtClient } from '@spectrtechnology/sdk';

const client = new SpectrtClient({ apiKey: process.env.SPECTR_API_KEY });

const result = await client.generate({
  url:      'https://yourapp.com',
  testType: 'ui',
  steps: [
    'Navigate to login page',
    'Enter valid credentials and submit',
    'Assert dashboard is visible',
  ],
});

console.log(result.code);      // Generated Playwright TS
console.log(result.evalScore); // deepEval quality gate score

Open GitHub Actions generator

deepEval Quality Gates

Every generated test is scored across five metrics. Tests below threshold are flagged before they reach your CI pipeline.

Faithfulness≥75

Test matches the user story intent

Coverage≥70

All described behaviour has assertions

Correctness≥75

Assertions are logically sound

AssertionDensity≥60

Enough verification per line of test

AntiPattern≥80

No sleeps, hardcoded selectors, empty catches

Test everything.Write nothing.

REST API

SDK & CI/CD

deepEval Quality Gates

Test everything.
Write nothing.