Changelog

Release history for KRINO Dokim

v1.5.1, May 2026

Security & Reliability

• Comprehensive security hardening across desktop app and authentication
• Improved input validation on all API endpoints
• Strengthened update integrity verification
• Enhanced error handling to prevent information disclosure in logs

Bug Fixes

• Fixed startup failure on all platforms: the desktop app crashed on launch due to a missing NumPy internal module in the packaged build
• Fixed the Retry button on the startup error screen: it now properly kills the stale server process and retries startup rather than doing nothing
• Fixed the Copy Error button on the startup error screen: clipboard access was blocked in the desktop context and silently failed
• Server crash output is now captured to a log file, making startup failures diagnosable
• Increased startup timeout from 30 to 60 seconds to accommodate first-run antivirus scans on Windows

v1.5.0, April 2026

Specification Linting

• KRINO Dokim™ now analyzes your acceptance criteria before generating test cases
• Flags vague language, untestable conditions, missing actors, and ambiguous outcomes
• Catching specification issues early leads to better test cases and fewer corrections after the fact
• Linting runs on demand: use it when you want it, skip it when you don't

Output Critique

• After generation, run a review pass over the produced test cases
• Identifies gaps such as missing negative scenarios, missing edge cases, and under-specified expected results
• Each finding includes a concrete suggestion for how to address it

Reference Examples

• Provide your own example test cases to guide how KRINO Dokim formats and structures its output
• Useful for matching house style conventions or aligning with an existing test suite
• Examples are saved with the story and persist across sessions and batch checkpoints

Batch File Preview

• Sort any column by clicking its header
• Filter rows by typing in the search box
• Select or deselect rows using checkboxes to control which stories are included in the batch run

Results & Review

• Review section now appears above the export section, keeping actions closer to the generated output

AI Providers & Pricing

• Gemini model pricing updated to reflect April 2026 changes
• New Gemma 4 models added to the Gemini provider
• Added Gemini Deep Research models
• Discontinued preview models removed

Reliability

• Fixed a regression where Modify and Rerun buttons on other stories were not correctly disabled during processing

v1.4.0, March 2026

Subscriptions & Billing

• Added subscription activation panel with trial countdown and expiry warning
• Free trial with full access: card collected at sign-up, not charged until trial ends
• Monthly and Annual Pro plans available
• Manage your subscription in Settings: upgrade, cancel, or view invoices without leaving the app
• Offline support: app continues working without an internet connection

Account & Authentication

• Email verification required before signing in
• Forgot password flow: reset via email link from the sign-in screen
• Device management in Settings: view registered devices and remove unused ones
• Sign out blocked while generation or batch processing is active
• Account lockout protection: progressive cooldown after repeated failed login attempts
• Password reset automatically unlocks your account and signs out all devices

Desktop UX

• First-run welcome card: guides new users through provider setup on first launch
• Auto-detects local AI providers on startup
• Form state preserved across sessions: story fields restore on relaunch
• Provider and coverage level selections remembered between sessions
• “New Story” resets the form without reloading: previous results are preserved
• System dialogs replaced with native in-app toast notifications
• Timeout warning elevated to a toast: visible without scrolling
• Batch step indicator shows current story number and total during processing
• Cancel button visible as soon as generation starts
• Test results rendered as formatted BDD cards
• Keyboard shortcuts: Ctrl+Enter (submit), Ctrl+1/2 (tabs), Ctrl+, (settings), Space (pause/resume batch), Esc (close panel)
• Redesigned About panel with version info and quick links
• Accessibility improvements across panels and form controls

AI Providers & Pricing

• Model pricing updated across all providers: verified against live APIs
• Free models now clearly labeled in the model dropdown
• 8 free models available via Google Gemini: including Gemma 3 and Gemini Flash previews
• Added support for latest OpenAI GPT-4.1 models and Gemini 3.1 series
• Removed deprecated models that are no longer available from providers

Installation & Updates

• Windows installer: run the setup wizard instead of extracting a ZIP
• macOS DMG: drag to Applications to install
• Linux DEB package: install with a single dpkg command
• In-app update notifications: a banner appears when a new version is available
• One-click update download: download and install without leaving the app
• Downloads are verified for integrity before installation
• Silent update install on Windows: app restarts automatically after update
• Welcome email sent after account verification with a quick-start guide

Security & Reliability

• Comprehensive security hardening across authentication, file handling, and error reporting
• Error messages no longer expose internal details: clear, actionable messages only
• Stolen session tokens are automatically detected and revoked
• File operations validated to prevent unauthorized directory access
• Old processing data automatically cleaned up on startup
• Improved server responsiveness under concurrent load

v1.3.0, January 2026

Desktop Application

• KRINO Dokim is now a standalone desktop app: no browser or terminal required
• Native installers for Windows, macOS (Apple Silicon + Intel), and Linux
• Splash screen with loading indicator and startup status
• Native file save dialogs for all exports: files save directly to your chosen location
• Single-instance lock prevents multiple copies from running simultaneously

Cloud AI Providers (BYOK)

• Added OpenAI, Anthropic, and Google Gemini as cloud AI providers
• Bring Your Own Key (BYOK): API keys stored locally, never sent to KRINO Dokim servers
• Dynamic model discovery: available models fetched from each provider's API
• Free models highlighted in the dropdown for easy identification
• Real-time token usage and cost tracking per story and per batch

Processing Reliability

• Crash recovery: resume interrupted batch runs on relaunch
• Automatic retry for transient failures
• Pause and cancel preserve your progress: no lost work on resume
• Automatic reconnection after network interruptions
• Truncation detection warns when an AI response is cut short

Coverage Levels

• Added Quick, Standard, and Exhaustive coverage presets
• Coverage level selector available in both single story and batch modes
• AI reasoning validates generated test cases against the selected coverage level

Batch Processing

• CSV file support added alongside Excel (.xlsx, .xls)
• File preview step before processing: review row count and column mapping
• Rerun failed stories individually or all at once
• Modify and regenerate individual stories without reprocessing the entire batch
• Cost dashboard tracks cumulative spend across a batch run
• Export to Excel, JSON, CSV, Jira Xray, and TestRail

v1.2.0, December 2025

Cloud API Integration

• Added OpenAI, Anthropic, and Google Gemini as cloud AI providers
• API key validation with immediate health check on save
• Dynamic model list fetched from each provider
• Token usage and cost tracking displayed in run metrics
• API keys never exposed in error messages or logs

Coverage Levels

• Added coverage level selector: Quick, Standard, and Exhaustive
• Coverage level included in batch processing and export metadata

Batch Enhancements

• Rerun individual failed stories without reprocessing the entire batch
• Modify story details inline and regenerate test cases
• Collapsible settings panel per story in batch results

Error Handling

• Improved error messages with actionable guidance for common issues
• File upload validation strengthened

v1.1.0, November 2025

Stability & Resilience

• Pause/resume timer preserves elapsed time across cycles
• ETA calculation with confidence indicator based on stories processed
• Contextual error messages: network, timeout, and connection failures each get specific guidance

Retry Logic

• Automatic retry for transient failures
• Retries respect pause and cancel state

UI Improvements

• Unified metrics layout across single story and batch results
• Inline time metrics: elapsed time and ETA with confidence indicator

v1.0.0, October 2025

Initial Release

• AI-powered BDD test case generation from user stories
• Supports Ollama and LM Studio as local AI providers
• Single story mode: enter a user story, generate Gherkin test cases, copy or export
• Batch processing mode: upload an Excel file with multiple stories
• Real-time log streaming with color-coded log levels
• Progress bar with percentage and story count for batch runs
• Pause, resume, and cancel controls for batch processing
• Export results to Excel

Contact

Found a bug or have a feature request? Email support@krinosystems.com.