PrimeCompassPrimeCompass

PrimeCompass Documentation

Everything you need to know about using PrimeCompass to map your application, detect drift, and ship with confidence.

Overview

PrimeCompass is an intelligent application testing and monitoring platform that continuously understands how your application behaves, tracks changes over time, and gives teams early, objective insight into risk and release readiness.

The platform automatically maps your entire application, discovers user journeys, creates test scenarios, detects changes between releases, and provides AI-powered insights — all without manual test scripting.

How It Works

1

Configure

Point PrimeCompass at your application URL, define user personas, and select the target environment.

2

Discover

PrimeCompass automatically explores your application, mapping pages, forms, workflows, and interactions.

3

Baseline

Capture a snapshot of your application's current behaviour as a baseline — the expected state.

4

Compare

Run new discoveries against your baseline. PrimeCompass identifies drift, breaking changes, and risk areas.

Dashboard

The Dashboard is your central hub for managing all applications. Each application card displays its current status, key metrics, and quick actions.

Application Status

New

Application registered but no discovery has been run yet.

Discovered

At least one discovery run has completed.

Baselined

A baseline has been created, ready for comparison.

Metrics at a Glance

Each application card shows the number of discoveries, baselines, scenarios, and personas — giving you an instant picture of coverage and activity.

Discovery

Discovery is PrimeCompass's core capability. It runs an intelligent, AI-guided crawler that explores your application, identifies pages, forms, interactions, and workflows, then generates executable test scenarios automatically.

Setting Up a Discovery

1

Target URL & Environment

Enter your application's URL and select the environment (Development, Staging, Production, or Local). PrimeCompass remembers your previous configuration.

2

Personas

Define user roles (e.g. Admin, Employee, Customer) with their login credentials. PrimeCompass logs in as each persona and discovers role-specific workflows.

3

Devices & Browsers

Choose target devices (Desktop, Tablet, Mobile) and browsers (Chrome, Firefox, Safari) for cross-platform coverage.

4

Discovery Modes

Run personas in Parallel (simultaneous, faster) or Sequential (one at a time, controlled). Configure the maximum concurrent persona count.

Discovery Results

After a discovery completes, the Overview page provides a comprehensive breakdown:

  • All discovered pages with types (Form, List, Detail, Navigation, etc.)
  • Scenarios grouped by business module
  • Step-by-step actions with pre/post screenshots
  • Entity data created or modified during exploration
  • AI-generated insights and suggested follow-up actions

Baselines

A Baseline is a snapshot of your application's expected behaviour — the set of scenarios, workflows, and structures that represent the "known good" state.

Creating a Baseline

Create a new baseline with a name and description, then populate it by importing scenarios from discovery runs, recordings, the composer, or external files.

Baseline Detail

Modules & Scenarios

View all scenarios organized by business module. Expand any scenario to see its steps, BDD feature syntax, Playwright scripts, and screenshots.

Import Sources

Track where each scenario came from — which discovery, recording, or file import. Reimport from sources to update scenarios.

Lifecycle Management

Archive old baselines, supersede with new versions, or fork an existing baseline to branch your testing.

Comparison & Drift Detection

Drift detection compares your baseline (expected state) against a new discovery run to identify what has changed between releases.

How Comparison Works

Select a baseline and a discovery run, then start the comparison. PrimeCompass uses a multi-pass matching engine to align scenarios and identify differences.

Comparison Results

Exact Matches

Scenarios identical in both versions — no action needed.

Partial Matches

Scenarios with minor differences that may need review.

New Scenarios

Workflows present in the new version but absent from the baseline.

Missing Scenarios

Baseline scenarios not found in the new version — potential regressions.

AI-Generated Insights

The comparison overview includes AI-powered insight cards that highlight the most impactful changes, risk areas, and recommended actions — with links to drill down into details.

Recorder

The Recorder captures your manual interactions with your application and converts them into executable test scenarios. Use it when you want to capture specific workflows that automated discovery may not reach.

Recording a Workflow

1

Enter URL & Environment

Enter your application URL and select the environment. Optionally assign a persona for the recording session.

2

Start Recording

Click Start Recording — PrimeCompass opens your application in a browser. Interact with your application naturally.

3

Interact

Every click, form fill, navigation, and selection is captured automatically.

4

Stop & Export

Stop the recording when your workflow is complete.

Recording Output

Each recording generates a timeline of actions with screenshots. Export in multiple formats:

Human-readable scenariosPlaywright scriptsBDD feature files

Scenario Composer

The Composer is a visual tool for designing complex test workflows by arranging scenarios into parallel lanes.

Module Browser

Browse all discovered modules and scenarios in the left panel. Filter by name, module, or status.

Lanes

Drag scenarios into lanes on the canvas. Scenarios within a lane run sequentially; different lanes run in parallel.

Execute & Review

Save your composition and run it. Track progress per lane, view pass/fail results, screenshots, and logs.

CompAssist AI

CompAssist is PrimeCompass's built-in AI assistant, available from any page via the chat panel on the right side of the screen.

Capabilities

  • Ask natural-language questions about your discoveries, baselines, and comparisons
  • Generate charts and visualizations on demand
  • Get summaries and risk assessments for discovery runs
  • Export answers as Excel, CSV, or PDF reports
  • Context-aware prompts that adapt to the page you're on

Example Prompts

“Show me the navigation graph”
“Summarize this discovery run”
“What areas have test gaps?”
“Explain these differences”
“What's the risk for this release?”

Test Artifact Generation

PrimeCompass automatically generates multiple test artifact formats from discovered scenarios, ready for integration into your CI/CD pipeline.

Generated Formats

.spec.ts

Playwright Scripts

Ready-to-run TypeScript test scripts with page objects, assertions, screenshots, and retry logic.

.feature

BDD Feature Files

Gherkin-format scenarios compatible with Cucumber, SpecFlow, and other BDD frameworks.

.json

Scenario Matrix

Structured JSON with scenario metadata, steps, test data, and cross-references.

.csv

CSV / TSV Reports

Spreadsheet-friendly exports for review, requirement tracing, and test management tools.

CI/CD Integration

Download generated scripts as a ZIP archive and commit them to your repository. Run them with any standard test runner:

Playwright TestJestGitHub ActionsGitLab CIJenkinsAzure Pipelines

Personas & Role-Based Testing

Personas represent different user roles that interact with your application. PrimeCompass uses personas to discover role-specific workflows and validate access controls.

Multi-Persona Discovery

  • Log in as each persona separately and map their unique views
  • Identify pages and features accessible to one role but not another
  • Discover approval workflows, admin-only functions, and permission boundaries
  • Generate per-persona scenario sets for targeted regression testing

Role Comparison

Compare what different personas can see and do. Identify access gaps, shared workflows, and role-specific scenarios to ensure your permission model works as intended.

Import & Export

Import Sources

Discovery runs

Import all or selected scenarios with preview

Recordings

Convert manual recordings into baseline scenarios

Composed workflows

Import saved compositions

External files

CSV, Excel, JSON, BDD feature files, or Playwright scripts

Long-running imports are processed in the background with progress tracking, notifications on completion, and the ability to reimport from any source.

Export Options

  • Download all test scripts as a ZIP archive
  • Export comparison reports as CSV or Excel
  • Export scenarios by module, persona, or format
  • Export system configuration as JSON for backup or team sharing

Settings

Configure PrimeCompass to match your team's workflow and infrastructure.

Configuration Categories

Discovery

Timeouts, crawl depth, concurrent browsers, form exploration, screenshot capture.

Execution

Test timeouts, retry policies, screenshot-on-failure, video recording.

Reporting

Report formats, email notifications, data retention policies.

Advanced

LLM provider for AI features, webhook integrations, API configuration.

Search any setting by keyword, export your configuration as JSON, or reset individual settings to their defaults.

Credits & Billing

Credits are the universal currency for every PrimeCompass operation. Each action consumes a deterministic, complexity-aware amount of credits — your tier sets the monthly budget, not the unit price. Heavier work costs more; everyone pays the same rate per op.

Tier allowances

Tier
Monthly credits
Discoveries / month
Add-on pack
Starter
100,000
5
20,000 / $12
Pro
500,000
25
50,000 / $39
Business
2,000,000
50
200,000 / $99
Enterprise
Unlimited
Unlimited
—

Tier allowance and discovery quota are designed to converge at average workload. A 1-persona × 200-page × standard discovery costs 20,000 credits — exactly 1/5 of Starter's 100k budget for 5 monthly runs, and 1/25 of Pro's 500k for 25 runs. Heavy users hit the credit ceiling first; light users have surplus credits for chat, comparisons, and scripts.

Cost formulas per operation

All costs derive from a base unit of 5,000 credits, scaled by the real complexity factors of each operation. Same rate across every tier.

Discovery run5,000 × personas × ⌈targetPages / 50⌉ × mode

mode = 1.0 standard · 2.5 deep AI · default targetPages = 100 if unspecified.

Example: 1 persona × 200 pages × standard = 20,000 credits.

Composition execute2,500 + 500 × scenarios

Visual mode applies a 1.5× multiplier on the total. Re-priced 50% in 2026-Q2.

Example: 10 scenarios standard = 7,500 credits; visual = 11,250.

Recording (start)2,500 + 500 × duration_minutes

Charged when you press Record; replay and convert are charged separately. Re-priced 50% in 2026-Q2.

Example: 5-minute recording = 5,000 credits; 30-minute = 17,500.

Recording replayFlat — 500 credits

Lightweight playback of an already-captured stream.

Example: Any replay = 500 credits.

Recording → script convertFlat — 2,000 credits

Converts a recording into a Playwright/Cypress script.

Example: Any conversion = 2,000 credits.

Baseline run50 × scenarios

Charged when you execute a baseline. Cost is finalized post-run from the actual scenarios played; the gate reserves a single-scenario amount up front.

Example: 20-scenario baseline = 1,000 credits.

Baseline comparison5,000 standard · 15,000 deep

Deep comparison uses LLM-driven semantic compare for nuanced drift detection. Re-priced 50% in 2026-Q2.

Example: Standard = 5,000; deep = 15,000.

CompAssist AI chat messageFlat — 200 credits

Per chat message. Sized so even heavy chat use stays within tier budget.

Example: Any message = 200 credits.

Script generation1,500 × scenarios

Single = one scenario; batch scales linearly with the number of scenarioIds.

Example: Single = 1,500; 10-scenario batch = 15,000.

Baseline / FSD importFlat — 3,000 credits

Importing an FSD/BRD as a baseline.

Example: Any import = 3,000 credits.

Credit lifecycle

1

Granted on subscription activation

Your tier's monthly allowance is credited to your account. The counter resets at the start of each calendar month.

2

Granted on credit-pack purchase

Top-up credits land in a separate purchased bucket valid for 90 days. They stack on top of your monthly allowance.

3

Deducted on every successful operation

The credit gate computes the cost from the request, reserves it before the handler runs, and commits the deduction when the response succeeds. Failed requests (4xx / 5xx) cost nothing.

4

Reset at the start of each calendar month

Monthly credits refill automatically. Purchased and bonus credits roll over (subject to their own expiry rules).

5

Top up anytime with a credit pack

From the Billing page in the app, buy your tier's pack (Starter $12 · Pro $39 · Business $99). Pack credits stack on top of your monthly allowance and are valid for 90 days.

How discovery starts are gated

Every POST /apps/:appId/discoveries call passes through two sequential checks:

  1. Discovery quota — counts your discovery runs this calendar month against your tier's maxDiscoveriesPerMonth. Blocks if you've already hit the cap.
  2. Credit balance — computes the cost of this specific run (using the formula above) and blocks with HTTP 402 if your remaining credits are short.

Whichever gate fires first wins. A heavy user runs out of credits before hitting their quota; a light user might never hit either; an average user converges on both at the same time.

Bring your own LLM (BYOK)

All tiers can configure their own LLM provider keys (Gemini, OpenAI, Anthropic, DeepSeek, or self-hosted Ollama) underSettings → LLM. Your key is used for outbound LLM calls during discovery, BDD generation, and chat — your own provider quotas and costs apply to those calls. PrimeCompass platform credits still meter the PrimeCompass-side work (queue management, browsers, storage, etc.).

Leaving a provider's key empty falls back to the platform default — usage counts toward your monthly plan credits as usual. Customer-side setting reads never expose the platform's default key value.

Invoices & receipts

Every successful subscription charge or credit-pack purchase generates a PDF tax invoice automatically. You can download invoices fromBilling → Invoices in the app, or wait for the email copy that arrives at your billing address within minutes of payment. For Indian customers, GSTIN can be added underBilling → Tax info; invoices then carry CGST + SGST or IGST depending on your state versus the seller's.

Need a copy of a missed invoice? Email support@primethoughts.com with your org name and the approximate date — we'll resend.