Changelog
What's new in PromptOT
February 2026v1.4.0
Editor Redesign & Error Tracking
- Redesigned prompt editor with dual-scroll layout and fixed header
- Card-based block layout with polished panels and visual cues
- Sentry error tracking for frontend and API with source map uploads
- SEO glossary and comparison pages for AI and prompt engineering terms
- Auto-polling for running evaluation runs with live status updates
February 2026v1.3.0
Evaluations UI Improvements
- Test case names displayed in evaluation results instead of truncated IDs
- Clickable test case names open a detail modal with input, expected output, variables, and criteria
- New evaluation result viewer modal showing actual LLM output, test input, and expected output side by side
- Model stats (latency, token breakdown) shown in result detail view
- Friendly model display names throughout the evaluations page
- Fixed evaluation run creation failing when no specific test cases were selected
February 2026v1.2.0
Member Role Management & Co-pilot Test Case Generation
- Generate test cases directly from the co-pilot — describe what to test and get domain-aware proposals
- Balanced test mix: happy path, edge cases, boundary tests, and adversarial inputs
- Per-test-case Accept/Reject with bulk Accept All / Reject All support
- Variable-aware generation exercises different variable combinations from your prompt
- Each test case includes 2-3 verifiable evaluation criteria (contains, regex, similarity)
- Organization owners can now change member roles directly from the settings page
- Inline role selector dropdown for quick role changes between owner, editor, and viewer
- Automatic project access adjustment when promoting a member to owner
February 2026v1.1.0
Prompt Co-pilot
- Conversational AI co-pilot for prompt editing — describe changes in natural language
- Multi-block actions: edit, add, remove, reorder, and toggle blocks from chat
- Inline diff previews with per-action Accept/Reject controls
- Prompt quality analysis with scoring, strengths, weaknesses, and suggestions
- Persistent conversation history per prompt with model selection
- Prompt engineering best practices built into the co-pilot's knowledge
February 2026v1.0.0-beta
Public Beta Launch
- Launched PromptOT to the public
- Landing page with full product overview
- Documentation and API reference
February 2026v0.9.0
Access Control & Collaboration
- Project-level access control for organization members
- Role-based permissions (owner, admin, member, viewer)
- Member invitation system
January 2026v0.8.0
AI-Powered Prompt Editing
- AI block rewriting (improve, expand, simplify, custom modes)
- Playground evaluation with OpenRouter integration
- Prompt usage documentation with code snippets
January 2026v0.7.0
Webhook Notifications
- Webhook delivery for prompt events (published, updated, deleted)
- HMAC-SHA256 payload signatures
- Exponential backoff retry (up to 5 attempts)
December 2025v0.6.0
API Delivery System
- Public API for prompt delivery (/api/v1/*)
- API key authentication with SHA-256 hashing
- Environment scoping (production vs development)
November 2025v0.5.0
Block-Based Composition
- Typed prompt blocks (role, context, instructions, guardrails, output_format, custom)
- Drag-and-drop block reordering
- Variable interpolation with {{key}} syntax
- Prompt compilation engine