PrimeCompass

PrimeCompass is an AI-powered application testing and quality intelligence platform built by PrimeThoughts Innovation. It autonomously crawls web applications, maps every page and workflow, and generates executable test scenarios — all without requiring source code access, browser extensions, or SDK integration. The platform is available as a cloud-hosted SaaS and targets engineering teams that want to catch test coverage gaps before they reach production.

What It Is

PrimeCompass sits in the automated testing and quality intelligence category. Its core job is to replace manual test discovery and scenario authoring with an AI-guided crawler that navigates a live application the way a real user would — handling authentication, single-page app routing, dynamic content, and multi-step flows. The output is a structured knowledge graph of the application, plus ready-to-run BDD feature files and Playwright/Cypress scripts. The platform vendor claims teams can go from a bare URL to a full test suite in under two hours, and that AI discovery surfaces 30–40% more pages than manual mapping.

Core Capabilities

• Autonomous discovery: AI crawler maps pages, forms, modals, and interactive elements; builds a knowledge graph with nodes and edges representing application structure.
• Multi-persona testing: Authenticates simultaneously as Admin, Editor, Viewer, Guest, or any custom role to surface permission gaps and RBAC issues.
• BDD scenario generation: Produces Given-When-Then Gherkin scenarios from real observed behavior; exports to Playwright, Cypress, and standard Cucumber.
• Baseline comparison and drift detection: Imports a Functional Specification Document (FSD) or a previous discovery run as a baseline, then diffs against new runs to flag added, removed, or changed pages and flows.
• Visual replay: Live VNC streaming during discovery and step-by-step action replay with timestamps, screenshots, and AI reasoning at each interaction.
• Compassist AI chat: Natural-language interface for querying the application knowledge graph — e.g., "Which pages require auth?" or "What changed since last week?"

Integration and Export Ecosystem

PrimeCompass exports generated test artifacts as .spec.ts Playwright scripts, .feature Gherkin files, .json scenario matrices, and .csv reports. CI/CD integrations cover GitHub Actions, GitLab CI, Jenkins, CircleCI, and Azure Pipelines. Issue tracker and notification integrations include Jira, Linear, Slack, and Microsoft Teams. Observability hooks connect to Datadog, Grafana, and PagerDuty. The platform also supports bring-your-own LLM keys (Gemini, OpenAI, Anthropic, DeepSeek, or self-hosted Ollama) for customers who want control over model selection and costs.

Credit-Based Usage Model

Every operation in PrimeCompass consumes credits from a monthly tier allowance. Discovery runs scale with the number of personas, pages crawled, and AI mode (standard vs. Deep AI at 2.5× multiplier). Other operations — baseline comparisons, script generation, Compassist chat messages, recordings — each have deterministic flat or formula-based costs. Failed requests (4xx/5xx) consume no credits. Purchased credit packs stack on top of monthly allowances and are valid for 90 days. The docs note that composition execute, recording, and baseline comparison costs were re-priced 50% downward in 2026-Q2.

Deployment and Security Model

PrimeCompass runs discovery in isolated, ephemeral browser containers. Discovery data is encrypted at rest and in transit. The vendor states the platform follows SOC2 compliance standards and maintains a 99.9% uptime SLA. Enterprise customers can choose data residency in US or EU regions and bring their own LLM API keys. No local installation, browser extension, or code change is required on the target application side.