v2.1.0

What's Changed

• Add POS highlighting feature flag and update documentation by @taruma in #31

Full Changelog: v2.0.0...v2.1.0

FrameLab v2.0.0: From Single Uploads to Full Multi‑Media Control

🎉 FrameLab v2.0.0

This is a major workflow and UX upgrade for FrameLab. v2.0 introduces flexible input modes (including text-only), true multi-media handling with editable tagging, richer editing controls, and stronger transparency/testing/documentation support.

✨ Highlights

• Text-only mode is now supported in both Primary Analysis and Refinement Loop (media is optional).
• Multi-media uploads are now supported in both phases (single or multiple image/video files).
• Per-item media tagging (@imageN / @videoN) is now editable and used in tagged payload composition.
• Added a dedicated Manage media tags dialog with full-size preview + tag editing.
• Added output editing dialogs for both phases, with edited outputs persisted into conversation context.
• Added a larger sidebar dialog editor for the System Prompt.
• Added optional session request logging with JSON export.
• Added configurable hero notices via config.toml.
• CI/testing improvements: offline pytest coverage + opt-in live smoke execution with --live and .env loading.

🔄 Changes

• Workflow terminology updated to:
  • Primary Analysis (formerly Phase 1)
  • Refinement Loop (formerly Phase 2)
• System Prompt behavior updated:
  • Editable textbox is now the source of truth
  • Preset dropdown now applies via explicit Load action
• Request Transparency previews improved for multi-media payloads (including media summaries + media-tag mapping).
• Copy actions now support both Copy Plain Text and Copy Markdown.
• Phase headers and reasoning sections updated with Material icon labels.
• In-app notices refined (including Xiaomi/OpenRouter update messaging).

🛠️ Fixes

• Improved media-tag persistence when media items are added/removed (signature-based mapping).
• Improved setup reliability by bundling en_core_web_sm via dependency wheel and aligning docs around uv sync.
• Improved editing ergonomics with wider system prompt and phase-output edit dialogs.

📚 Documentation

Updated:

• README.md
• AGENTS.md
• docs/REFERENCE.md
• docs/TESTING.md

What's Changed

• Add offline CI pytest suite and improve testing documentation by @taruma in #17
• Add configurable hero notices and update messaging in UI by @taruma in #18
• feat(ui): make media uploads optional in both analysis phases by @taruma in #20
• Add options for copying plain text and markdown outputs by @taruma in #21
• Refactor UI terminology and enhance phase headers with icons by @taruma in #23
• fix(deps): bundle spaCy model wheel and sync guidance by @taruma in #24
• Add editable dialogs for Phase outputs and refactor UI components by @taruma in #25
• Add tagged multi-media uploads and improve tag management by @taruma in #26
• Enhance System Prompt UI with editable dialog and improved layout by @taruma in #28
• feat: add optional session request logging and JSON export by @taruma in #29
• V2.0 by @taruma in #19

Full Changelog: v1.2.1...v2.0.0

v1.2.1

Fixed

• Fixed POS highlighting so enabling it no longer breaks markdown rendering in model outputs.
• Updated highlighter behavior to be markdown-aware by preserving fenced code blocks, inline markdown spans, and line-prefix structure markers while highlighting plain text tokens.

Full Changelog: v1.2.0...v1.2.1

v1.2.0: Video Support & Prompt Management

This release expands FrameLab’s multimodal workflow with video support and improves prompt management for faster iteration.

✨ Highlights

• Added support for image + MP4 video inputs in both Phase 1 and Phase 2. Fix #7
• Added folder-based prompt presets with optional metadata sidecars. Fix #3
• Added a new initial preset: video technical hybrid script.

🔧 Improvements

• Removed leftover debug reasoning-effort caption from chat rendering. Fix #6
• Updated copy and docs to use broader “video” terminology (instead of MP4-only wording where appropriate).

📚 Documentation

• Expanded AGENTS.md guidance for video workflows and prompt preset behavior.
• Clarified optional spaCy POS-highlighting requirements. Fix #4

✅ Notes

• No breaking changes.
• Existing setup flow remains the same (uv run run.py).

What's Changed

• V1.2: Video Support & Prompt Management by @taruma in #8

Full Changelog: v1.1.1...v1.2.0

v1.1.1

See release v1.1 here: https://github.com/taruma/framelab/releases/tag/v1.1.0

Patch Fix

• Made Phase 1 and Phase 2 prompt fields editable in the UI so users can tune instructions per run without leaving the workflow.
• Cleaned up request transparency preview presentation for a clearer, more compact phase action experience.

v1.1.0

This release focuses on usability and configuration improvements to make FrameLab easier to set up and safer to operate during analysis runs.

✨ Highlights

• Added provider presets via config.toml for faster endpoint/model switching.
• Added unified API key env fallback with LLM_API_KEY support (plus provider-specific compatibility).
• Added a new hero landing section with branding, workflow guidance, badges, and creator credit.
• Added per-phase Request Transparency previews so users can inspect request metadata/payload before running.
• Improved request safety with UI locking during active processing to prevent duplicate submissions.

🔧 Improvements

• API Setup is now wrapped in an expander for cleaner layout.
• Transparency previews are now rendered inline with each phase action panel for clearer workflow context.
• Setup/documentation were updated to match the provider preset + key fallback flow.
• Added requirements.txt for pip-based installs and set Poetry to dependency-management-only (package-mode = false).

🐛 Fixes

• Improved phase layout consistency by removing placeholder-based transparency rendering and placing action buttons after each inline preview.

✅ Compatibility

• No breaking changes in this release.
• Existing uv run run.py workflow remains unchanged.

What's Changed

• V1.1.0 by @taruma in #1

New Contributors

• @taruma made their first contribution in #1

Full Changelog: v1.0.0...v1.1.0

FrameLab v1.0.0

Lightweight Multimodal AI for Cinematic Image Analysis

We're excited to announce the first release of FrameLab! 🎉

What is FrameLab?

FrameLab is a Streamlit-powered web app that brings professional-grade cinematic image analysis to your browser. Upload a reference image, get detailed technical breakdowns on composition, lighting, and optics — then refine your analysis through a correction loop.

✨ Key Features

• Multi-Provider Support — Works with any OpenAI-compatible endpoint.
• Live Streaming — Watch the analysis generate in real-time
• Two-Phase Workflow — Initial analysis + correction loop for iterative refinement
• Reasoning Display — See the model's "thought process" expand in real-time
• Copy Made Easy — One-click plain-text copy for your results
• Zero Fuss — Just streamlit and openai as dependencies

🛠️ Quick Start

uv sync
uv run run.py