TECHNICAL_SPEC_0.2

QAi
ROIST AGENTIC OS

STATUS: ALPHA_CORE // SECURITY: INTEGRATED

Replacing passive test management with A UNIFIED COGNITIVE ARCHITECTURE.
Write. Run. Heal. Secure.

BETA RELEASE & BYOK REQUIREMENT

QAi Sovereign is currently operating as a Beta release. To guarantee zero data telemetry and local sovereignty, cognitive functions within the desktop application require you to Bring Your Own Key (BYOK). Users must provide their own AI API keys (e.g., DeepSeek, Gemini) for script generation and reasoning.

01 // CORE_MODULES
MOD_DB_01

DASHBOARD

The Command Center. High-performance metrics aggregator pulling real-time statistics from the Execution Engine. Features the GlobalStatusSystem and trend heatmaps.

MOD_TE_02

TEST EDITOR

The Architect. A dense, Neo-Brutalist IDE. Governs the TestCaseItem data model with CRUD operations and AI-assisted expansion of steps via the IAIService.

MOD_AS_03

AUTOMATION STUDIO

The Automator. The heavy-lifting execution core. Leverages a BrowserContextPool to lease Playwright instances. Translates natural language into JSON actions.

MOD_AT_04

API TESTING

The Communicator. High-speed REST interaction layer. Features an auto-token extraction system that detects authentication handshakes to automate subsequent requests.

MOD_SV_05

STORYSIGHT

The Visionary. Predictive analysis engine using a hybrid of LLM reasoning and ML-based risk scoring to find defects before the first line of code is written.

MOD_AH_06

AI HUB

The Oracle. Central routing point for AI interactions. Manages multiple providers (DeepSeek, ChatGPT, Gemini, Ollama) with seamless auto-failover and aggressive caching.

MOD_LT_07

LOAD TESTING

The Tank. Stress-testing module designed to simulate high-concurrency scenarios. Spawns hundreds of threads to measure the breaking point of infrastructure.

MOD_RT_08

SECURITY / RED TEAM

The Assassin. Active-exploit engine. Features a Red Team Agent for fuzzing (SQLi, XSS) and a SAST module for auditing local code for leaked credentials.

MOD_SH_09

SELF-HEALING

The Medic. Detects DOM changes during execution. Finds the new element using semantic similarity and updates the ObjectRepository.

MOD_VA_10

VALUE ANALYTICS

The Accountant. ROI calculation engine. Translates "Hours Saved" and "Defects Caught" into actual financial impact metrics and velocity trends.

MOD_MS_11

MILESTONES

The Project Manager. Bridge between high-level project goals and test cases. Manages ProjectItems to provide a kanban-style insight into release readiness.

MOD_SC_12

SOURCE CONTROL

The Archivist. Direct Git integration. Treats Test Cases as Code, allowing for commits, branches, and merges directly within the QA interface.

02 // RELIABILITY_ARCH
30s

ADAPTIVE_TIMEOUTS

Intelligent processing windows designed for complex reasoning and deep-dive analysis without timeouts.

INF

IDEMPOTENT_TASKS

Self-healing tasks capable of indefinite re-execution without state corruption or data loss.

ZERO

SILENT_FAILURES

Aggressive error detection providing immediate, actionable feedback upon any system discrepancy.

03 // QAi_STANDARD

SOVEREIGN AGENTIC SYSTEM

QAi operates behind your firewall. The OFFLINE BRAIN ensures neural weights remain behind your firewall. Data remains local. Access is controlled. Sovereignty is non-negotiable.

[SEC_ENCR: DPAPI] // [SEC_TAMPER: AUTO_SHUTDOWN]

INDUSTRIAL ENGINE PRECISION

Engineered for endurance. Our CIRCUIT BREAKER technology negates node downtime by routing intelligence through healthy nodes. MONKEY GUARD protects against erratic input and adversarial injection.

[PERF_THR: MULTI] // [CODE_VER: ROSLYN]

DEPLOY INTELLIGENCE

CLAIM ACCESS

VERSION 0.2 // WIN64 // SOVEREIGN EDITION