v3.2.0
Critical Code Studies Workbench v3.2.0
CCS Methodology v2.7 | Next.js 16 / React 19 / TypeScript 5
A web application for close reading and hermeneutic analysis of software as cultural artefact.
What's in this release
Architecture refactoring: Extracted WorkbenchLayout.tsx (4,748 → 1,163 lines, 75% reduction) and conversation/page.tsx (2,191 → 155 lines, 93% reduction) into 6 custom hooks and 3 components. Pure structural refactor with zero logic changes. Fixed pre-existing auto-save bug where stale file handles threw errors instead of recovering gracefully.
Highlights
- Three analysis modes: Analyze Code (critique), Learn Methods (interpret), Create Code (vibe coding)
- IDE-style workbench: Three-panel layout with file tree, code editor with inline annotations, and AI chat
- 16 sample projects spanning 1958–2026 (FLOW-MATIC, ELIZA, Apollo 11, Colossal Cave, GNU Emacs, Agrippa, Markdown, Git Stash, 10 PRINT, Transformer, Esoteric Languages, and more)
- Multi-provider AI: Anthropic Claude, OpenAI, Google Gemini, Ollama (local)
- Cloud collaboration: Supabase-powered project sharing with OAuth, real-time sync, invite links, member management
- 10 nostalgic skins: Atari 2600, BBC Micro, C64, ELIZA, Geocities, HyperCard, Myspace, Teams, Teletext, Vaporwave
- Six annotation types: Observation, Question, Metaphor, Pattern, Context, Critique
- AI auto-annotation: Sparkles button for AI-suggested annotations
- Threaded annotation replies with real-time sync
- Project library with submission, review, and accessioning workflow
- PWA support with offline capabilities
- Dark mode with six accent colour palettes
- Session log export in JSON, Text, and PDF formats
Getting started
git clone https://github.com/dmberry/CCS-WB.git
cd CCS-WB
npm install
npm run devOpen http://localhost:3000 and configure your AI provider in Settings.