@peaske7/readit 0.3.0-rc.0 → 0.3.0-rc.2

package/.claude/settings.json

@@ -1,18 +0,0 @@
-{
-  "project": {
-    "name": "readit",
-    "description": "CLI tool for reviewing Markdown documents with inline comments",
-    "version": "0.1.0"
-  },
-  "context": {
-    "entryPoints": ["src/cli.ts", "src/App.tsx"],
-    "documentation": ["CLAUDE.md", ".claude/user-stories.md"],
-    "testFile": "test.md"
-  },
-  "preferences": {
-    "language": "typescript",
-    "style": "functional",
-    "formatting": "prettier",
-    "imports": "esm"
-  }
-}

package/.claude/user-stories.md

@@ -1,333 +0,0 @@
-# User Stories
-
-## Core Workflow
-
-### US-001: View Document in Browser
-
-**As a** writer/reviewer
-**I want to** view my Markdown or HTML document rendered in a browser
-**So that** I can review the content in a more readable format
-
-**Acceptance Criteria:**
-
-- Run `readit document.md` or `readit document.html` and browser opens with rendered content
-- markdown-it for Markdown (server-side rendering with shiki syntax highlighting)
-- Clean light theme with good typography for reading
-
----
-
-### US-002: Add Comment to Selected Text
-
-**As a** reviewer
-**I want to** select text and add a comment
-**So that** I can note feedback or changes needed
-
-**Acceptance Criteria:**
-
-- Select any text in the document
-- Input area appears in margin to enter comment
-- ⌘+Enter saves the comment
-- Comment appears as margin note next to highlighted text
-- Selection is preserved as reference
-
----
-
-### US-003: Manage Comments
-
-**As a** reviewer
-**I want to** edit, delete, and copy individual comments
-**So that** I can refine my feedback
-
-**Acceptance Criteria:**
-
-- Each margin note shows: selected text (truncated) + comment text
-- Click to edit → inline textarea editing
-- Delete button → removes comment and highlight
-- Copy button → copies "Selected: ... Comment: ..." format
-
----
-
-### US-004: Export All Comments
-
-**As a** reviewer  
-**I want to** export all comments at once  
-**So that** I can use them with AI or share with others  
-
-**Acceptance Criteria:**
-
-- "Copy All" button copies formatted prompt to clipboard
-- "Export JSON" button downloads JSON file with all comments
-- Format is AI-friendly (clear structure)
-
----
-
-### US-005: Persistent Comments ✅
-
-**As a** reviewer
-**I want to** close and reopen the tool without losing comments
-**So that** I can continue review sessions
-
-**Acceptance Criteria:**
-
-- Comments saved to `.comments.md` files in `~/.readit/comments/`
-- Comments restored when reopening same file
-- Different files have separate comment storage
-- Human-readable markdown format for LLM/git compatibility
-
-**Status:** Implemented (file-based storage)
-
----
-
-## Planned Features
-
-### US-006: File-Based Comment Storage ✅
-
-**As a** reviewer
-**I want to** have my comments stored in a plain markdown file
-**So that** I can version them with git, share them, and use them with LLMs and other tools
-
-**Acceptance Criteria:**
-
-- Comments stored in `~/.readit/comments/{hashed-path}/{filename}.comments.md`
-- Format is human-readable markdown (selected text as quote, comment below)
-- ~~Existing localStorage comments migrated on first run~~ (localStorage removed)
-- Comments auto-save to file on add/edit/delete
-- ~~`--storage local|file` flag~~ (file-only, no flag needed)
-- File can be edited manually and changes reflected on refresh
-
-**Status:** Implemented
-
----
-
-### US-007: Comment Resolution Workflow
-
-**As a** reviewer
-**I want to** mark comments as resolved/unresolved and navigate between them
-**So that** I can systematically address feedback with LLM assistance
-
-**Acceptance Criteria:**
-
-- Each comment has `unresolved` or `resolved` status
-- New comments default to `unresolved`
-- Visual indicators: yellow = unresolved, green/muted = resolved
-- Navigate to next/previous unresolved comment (keyboard shortcuts)
-- Unresolved count displayed in header
-- Copy single comment with context for LLM discussion (⌘⇧C on margin note)
-- Toggle status with one click, optionally auto-navigate to next
-
-**Status:** Not implemented (planned for v0.3.0)
-
----
-
-### US-008: Highlight Commented Text ✅
-
-**As a** reviewer
-**I want to** see which parts of the document have comments
-**So that** I can quickly identify reviewed sections
-
-**Acceptance Criteria:**
-
-- Text with comments has yellow background highlight (`<mark>` tags)
-- Margin notes are positioned next to their corresponding highlights
-- Clicking highlight scrolls to margin note
-- Clicking quoted text in margin note scrolls to highlight
-
-**Status:** Implemented
-
----
-
-### US-009: Comment Categories
-
-**As a** reviewer  
-**I want to** categorize comments (TODO, Question, Suggestion, etc.)  
-**So that** I can organize and filter feedback  
-
-**Acceptance Criteria:**
-
-- Dropdown to select category when adding comment
-- Visual indicator (color/icon) per category
-- Filter sidebar by category
-- Export includes category information
-
-**Status:** Not implemented
-
----
-
-### US-010: Review Multiple Files ✅
-
-**As a** reviewer
-**I want to** review multiple Markdown files in one session
-**So that** I can review an entire document set
-
-**Acceptance Criteria:**
-
-- ~~`readit docs/*.md` opens multi-file view~~ `readit file1.md file2.md` or `readit <dir>` opens multi-file view
-- File tabs or sidebar navigation (TabBar.svelte)
-- Comments stored per file
-- Bulk export across all files
-
-**Status:** Implemented (multi-file args, directory scanning, `open` command, tab bar UI). Glob patterns not yet supported.
-
----
-
-### US-011: Collaborative Review
-
-**As a** team lead  
-**I want to** share review sessions with teammates  
-**So that** we can collaborate on document feedback  
-
-**Acceptance Criteria:**
-
-- Generate shareable link
-- Real-time sync of comments
-- Author attribution on comments
-- Conflict resolution for simultaneous edits
-
-**Status:** Not implemented (future consideration)
-
----
-
-### US-012: Export Document as PDF with Comments
-
-**As a** reviewer
-**I want to** export the current document view as a PDF
-**So that** I can share or archive the reviewed document with all highlights and comments visible
-
-**Acceptance Criteria:**
-
-- "Export PDF" button in the header
-- PDF includes the rendered document with highlighted text
-- Margin notes are included next to their corresponding highlights
-- PDF preserves the visual layout of the review interface
-- Filename follows pattern: `{original-filename}-reviewed.pdf`
-
-**Status:** Not implemented (future consideration)
-
----
-
-### US-013: Shell Integration with File Autocomplete ✅
-
-**As a** CLI user
-**I want to** autocomplete markdown file paths with `@<TAB>` in my shell
-**So that** I can quickly find and open documents without typing full paths
-
-**Acceptance Criteria:**
-
-- `readit @<TAB>` launches fzf picker for markdown files in cwd
-- `readit @partial<TAB>` pre-filters the picker with "partial" as query
-- Standard Tab completion works for subcommands and options
-- `readit completion zsh|bash|fish` outputs shell-specific setup scripts
-- `eval "$(readit completion zsh)"` installs everything in one line
-
-**Status:** Implemented (v0.7.0)
-
----
-
-### US-014: Neovim Plugin ✅
-
-**As a** Neovim user
-**I want to** render and review markdown documents directly from my editor
-**So that** I can stay in my editing workflow while reviewing
-
-**Acceptance Criteria:**
-
-- `:ReaditOpen` starts server and opens current buffer in browser
-- `:ReaditStop` stops the server
-- `:ReaditReload` saves buffer and triggers browser refresh
-- `:ReaditStatus` shows server info
-- Configurable keymaps (default `<leader>r` prefix)
-- Server auto-cleanup on VimLeavePre
-- `:checkhealth readit` verifies dependencies
-
-**Status:** Implemented (v0.7.0)
-
----
-
-### US-015: Live Reload on File Changes ✅
-
-**As a** reviewer
-**I want to** see the browser update automatically when I edit the source document
-**So that** I can see my changes reflected in real time
-
-**Acceptance Criteria:**
-
-- Browser updates within ~200ms of file save
-- Works with Vim/Neovim saves (write-to-temp-then-rename pattern)
-- SSE connection auto-reconnects if server restarts (exponential backoff)
-- Document HTML and comments fetched in parallel for fast reload
-
-**Status:** Implemented (enhanced in v0.7.0, base SSE existed since v0.1.0)
-
----
-
-## Technical Stories
-
-### TS-001: CLI Argument Parsing
-
-- Use Commander.js for CLI
-- Main command: `<file>`, `--port`, `--host`, `--no-open`, `--clean`
-- Subcommands: `list` (show all commented files), `show <file>` (display comments for a file)
-- Validate file exists and is Markdown (.md) or HTML (.html)
-- Graceful error messages
-
-### TS-002: Server Architecture
-
-- Bun.serve() serves API + static files
-- `/api/documents` list and add documents
-- `/api/document` fetch individual document HTML and headings
-- `/api/comments` CRUD operations for comments
-- `/api/settings` get/update user settings
-- `/api/heartbeat` SSE for browser connection tracking
-- `/api/document/stream` SSE for document updates (live reload)
-- Development mode proxies to Vite dev server
-- Production mode serves built assets from dist/
-
-### TS-003: Build System
-
-- Vite for frontend bundling
-- `bun build` for CLI bundling (target bun, ESM)
-- Single `bun run build` produces complete distributable
-- `bunx readit` works without global install
-
-### TS-004: File-Based Comment Storage
-
-- Store comments in `~/.readit/comments/{hashed-path}/{filename}.comments.md`
-- Markdown format for human/LLM readability:
-
-  ```markdown
-  ## Comment 1
-  > Selected text here
-
-  Comment text here
-
-  ---
-  ```
-
-- API endpoints: `GET/POST/PUT/DELETE /api/comments` for CRUD operations
-- Watch file for external changes (optional)
-
-### TS-005: Shell Integration Architecture
-
-- `shell/_readit` - Standard zsh compdef for subcommands, options, file args
-- `shell/readit.zsh` - Rich widget: `@` prefix triggers fzf markdown file picker
-- Uses `fd` for fast file listing, falls back to `find`
-- fzf with bat/head preview, initial query from text after `@`
-- `readit completion zsh|bash|fish` CLI subcommand outputs setup scripts
-- Bash completion via `complete -F`, Fish via `complete -c`
-
-### TS-006: Neovim Plugin Architecture
-
-- `nvim-readit/lua/readit/init.lua` - Core: setup, server management, commands, keymaps
-- `nvim-readit/plugin/readit.lua` - Auto-loader on markdown FileType
-- `nvim-readit/lua/readit/health.lua` - `:checkhealth` module
-- Discovers existing servers via `~/.readit/server.json`
-- Starts server with `jobstart()`, attaches files via `POST /api/documents`
-- Communicates with server via curl (non-blocking `jobstart`)
-
-### TS-007: File Watcher Resilience
-
-- `fs.watch()` with both `"change"` and `"rename"` event handling
-- On rename: retry loop (10 attempts, 200ms apart) to re-establish watcher
-- Debounced re-render (200ms) to avoid thrashing on rapid saves
-- SSE auto-reconnect with exponential backoff (1s -> 30s max)
-- Parallel fetch of document HTML + comments on reload

package/AGENTS.md

@@ -1,68 +0,0 @@
-# AGENTS.md
-
-Instructions for AI coding agents working on this project.
-
-## Project Context
-
-readit is a CLI tool for reviewing Markdown and HTML documents with inline comments. The workflow is:
-
-1. User runs `readit document.md`
-2. Server renders Markdown to HTML (markdown-it + shiki for syntax highlighting)
-3. Bun.serve() starts and opens browser with Svelte 5 frontend
-4. User selects text and adds comments
-5. Comments are saved as `.comments.md` files in `~/.readit/comments/`
-6. User can export comments for AI or apply back to Markdown
-
-## Development Commands
-
-```bash
-bun install           # Install dependencies
-bun dev               # Start development (CLI with --watch)
-bun run dev:client    # Start Vite dev server only
-bun run build         # Build for production (Vite + CLI)
-bun run test          # Run unit tests (Vitest)
-bun run test:e2e      # Run e2e tests (Playwright)
-bun run typecheck     # TypeScript check
-bun run check         # Biome lint + format check
-bun run check:fix     # Fix lint + format issues
-bun run format        # Format code with Biome
-```
-
-## Code Organization
-
-- `src/cli.ts` - CLI entry point (Commander.js)
-- `src/server.ts` - Bun.serve() server + API routes
-- `src/components/` - Svelte 5 UI components
-- `src/stores/` - Svelte 5 reactive stores (`.svelte.ts`)
-- `src/lib/` - Utility functions, rendering, i18n
-- `src/schema.ts` - TypeScript types
-- `shell/` - Zsh/Bash/Fish completions and `@` file picker widget
-- `nvim-readit/` - Neovim plugin (Lua)
-- `vscode-readit/` - VS Code extension
-
-## Testing
-
-- Unit tests are co-located with source files (`*.test.ts`)
-- E2E tests live in `e2e/` (Playwright)
-- Run `bun run test` before committing
-
-## Commit Guidelines
-
-- Use conventional commits (feat:, fix:, docs:, etc.)
-- Keep commits focused and atomic
-- Run `bun run typecheck && bun run check` before committing
-
-## Adding Features
-
-1. Check `.claude/user-stories.md` for planned features
-2. Update types in `src/schema.ts` first
-3. Create/update stores if needed (`src/stores/`)
-4. Create/update Svelte components (`src/components/`)
-5. Add tests
-
-## Key Files to Review
-
-- `CLAUDE.md` - Project overview and architecture
-- `.claude/user-stories.md` - User stories and acceptance criteria
-- `.claude/roadmap.md` - Development roadmap
-- `package.json` - Scripts and dependencies

package/Makefile

@@ -1,32 +0,0 @@
-.PHONY: dev build build-client build-server test test-client test-e2e clean
-
-# Development: Go server manages Vite child process
-dev:
-	cd go && go run ./cmd/readit -- --dev $(ARGS)
-
-# Production build: frontend first, then Go embeds it
-build: build-client build-server
-
-build-client:
-	bunx vite build
-
-build-server: build-client
-	rm -rf go/internal/server/dist
-	mkdir -p go/internal/server/dist
-	cp -r dist/. go/internal/server/dist/
-	cd go && go build -o ../dist/readit ./cmd/readit
-
-# Tests
-test:
-	cd go && go test ./...
-
-test-client:
-	bun run test
-
-test-e2e:
-	bun run test:e2e
-
-clean:
-	rm -rf dist go/internal/server/dist
-	mkdir -p go/internal/server/dist
-	touch go/internal/server/dist/.gitkeep

package/biome.json

@@ -1,79 +0,0 @@
-{
-  "$schema": "https://biomejs.dev/schemas/2.4.15/schema.json",
-  "vcs": {
-    "enabled": true,
-    "clientKind": "git",
-    "useIgnoreFile": true
-  },
-  "formatter": {
-    "enabled": true,
-    "indentWidth": 2,
-    "indentStyle": "space"
-  },
-  "assist": {
-    "enabled": true
-  },
-  "linter": {
-    "enabled": true,
-    "rules": {
-      "style": {
-        "noNonNullAssertion": "off"
-      },
-      "correctness": {},
-      "complexity": {
-        "noForEach": "off",
-        "noImportantStyles": "off"
-      },
-      "suspicious": {
-        "noArrayIndexKey": "warn",
-        "noExplicitAny": "warn",
-        "noUnknownAtRules": "off"
-      },
-      "a11y": {
-        "useKeyWithClickEvents": "off",
-        "noSvgWithoutTitle": "off",
-        "useSemanticElements": "warn",
-        "useFocusableInteractive": "warn"
-      },
-      "security": {}
-    }
-  },
-  "css": {
-    "parser": {
-      "tailwindDirectives": true
-    }
-  },
-  "files": {
-    "includes": [
-      "**",
-      "!!**/dist",
-      "!!**/build",
-      "!!**/node_modules",
-      "!!**/generated",
-      "!!**/prisma/generated",
-      "!!**/supabase/generated",
-      "!!**/.next",
-      "!!**/.turbo",
-      "!!**/.git",
-      "!!**/.temp",
-      "!!**/.wrangler",
-      "!!**/playwright-report",
-      "!!**/test-results"
-    ],
-    "ignoreUnknown": true
-  },
-  "overrides": [
-    {
-      "includes": ["**/*.svelte"],
-      "linter": {
-        "rules": {
-          "correctness": {
-            "noUnusedImports": "off",
-            "noUnusedVariables": "off",
-            "noUnusedFunctionParameters": "off"
-          }
-        }
-      }
-    }
-  ]
-}