claude-kanban 0.1.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md ADDED
@@ -0,0 +1,167 @@
1
+ # Claude Kanban
2
+
3
+ A visual Kanban board for AI-powered development with Claude. Run `npx claude-kanban` in any project directory to manage and execute AI coding tasks.
4
+
5
+ ## Features
6
+
7
+ - **Visual Kanban Board**: Drag-and-drop tasks across columns (Draft, Ready, In Progress, Completed)
8
+ - **AI Task Execution**: Click to run tasks with Claude Code CLI
9
+ - **Real-time Updates**: WebSocket streaming of execution output
10
+ - **Parallel Execution**: Run up to 3 tasks concurrently
11
+ - **AFK Mode**: Automated loop execution for hands-off development
12
+ - **AI Task Generation**: Describe what you want, get structured tasks
13
+ - **Task Templates**: Pre-built templates for common patterns
14
+
15
+ ## Installation
16
+
17
+ ```bash
18
+ # Run directly with npx (no installation needed)
19
+ npx claude-kanban
20
+
21
+ # Or install globally
22
+ npm install -g claude-kanban
23
+ claude-kanban
24
+ ```
25
+
26
+ ## Requirements
27
+
28
+ - Node.js 18+
29
+ - Claude Code CLI (`claude`) installed and configured
30
+
31
+ ## Usage
32
+
33
+ ### Quick Start
34
+
35
+ ```bash
36
+ cd your-project
37
+ npx claude-kanban
38
+ ```
39
+
40
+ This will:
41
+ 1. Create `.claude-kanban/` directory with config files
42
+ 2. Create `scripts/ralph.sh` and `scripts/ralph-once.sh`
43
+ 3. Start a local web server
44
+ 4. Open your browser to the Kanban board
45
+
46
+ ### CLI Options
47
+
48
+ ```
49
+ Usage: claude-kanban [options]
50
+
51
+ Options:
52
+ -V, --version output the version number
53
+ -p, --port <number> Port to run server on (default: "4242")
54
+ -n, --no-open Do not auto-open browser
55
+ --init Re-initialize project files
56
+ --reset Reset all tasks (keeps config)
57
+ -h, --help display help for command
58
+ ```
59
+
60
+ ### Creating Tasks
61
+
62
+ **Manual**: Click "+ New" to create a task with title, description, and verification steps.
63
+
64
+ **AI-Assisted**: Click "AI Create", describe what you want, and Claude generates structured task details.
65
+
66
+ **Templates**: Click "Templates" to use pre-built patterns for auth, CRUD, API endpoints, etc.
67
+
68
+ ### Running Tasks
69
+
70
+ 1. Move tasks from "Draft" to "Ready" when requirements are clear
71
+ 2. Click "Run" on a Ready task to execute with Claude
72
+ 3. Watch real-time output in the execution log
73
+ 4. Task automatically moves to "Completed" on success or shows "Failed" status
74
+
75
+ ### AFK Mode
76
+
77
+ Click "AFK" to run multiple iterations automatically:
78
+ - Set maximum iterations
79
+ - Set concurrent task limit (1-3)
80
+ - Agent picks up tasks from "Ready" column
81
+ - Continue until all tasks complete or max iterations reached
82
+
83
+ ### Shell Scripts
84
+
85
+ The tool generates shell scripts for CLI usage:
86
+
87
+ **ralph.sh** - Run multiple iterations:
88
+ ```bash
89
+ ./scripts/ralph.sh 10 # Run 10 iterations
90
+ ```
91
+
92
+ **ralph-once.sh** - Run a single iteration:
93
+ ```bash
94
+ ./scripts/ralph-once.sh # Pick highest priority task
95
+ ./scripts/ralph-once.sh task_abc123 # Run specific task
96
+ ```
97
+
98
+ ## Project Structure
99
+
100
+ ```
101
+ your-project/
102
+ ├── .claude-kanban/
103
+ │ ├── prd.json # Task list (Product Requirements Document)
104
+ │ ├── progress.txt # Agent memory/progress log
105
+ │ └── config.json # Configuration
106
+ └── scripts/
107
+ ├── ralph.sh # AFK automation script
108
+ └── ralph-once.sh # Single-run script
109
+ ```
110
+
111
+ ## Configuration
112
+
113
+ Edit `.claude-kanban/config.json` to customize:
114
+
115
+ ```json
116
+ {
117
+ "agent": {
118
+ "command": "claude",
119
+ "permissionMode": "acceptedits",
120
+ "model": null
121
+ },
122
+ "project": {
123
+ "testCommand": "npm test",
124
+ "typecheckCommand": "npm run typecheck",
125
+ "buildCommand": "npm run build"
126
+ },
127
+ "execution": {
128
+ "maxConcurrent": 3,
129
+ "timeout": 30
130
+ }
131
+ }
132
+ ```
133
+
134
+ ## API Reference
135
+
136
+ The server exposes a REST API:
137
+
138
+ | Method | Endpoint | Description |
139
+ |--------|----------|-------------|
140
+ | GET | `/api/tasks` | Get all tasks |
141
+ | POST | `/api/tasks` | Create task |
142
+ | POST | `/api/tasks/generate` | AI-generate task |
143
+ | PUT | `/api/tasks/:id` | Update task |
144
+ | DELETE | `/api/tasks/:id` | Delete task |
145
+ | POST | `/api/tasks/:id/run` | Run task |
146
+ | POST | `/api/tasks/:id/cancel` | Cancel task |
147
+ | GET | `/api/templates` | Get templates |
148
+ | POST | `/api/afk/start` | Start AFK mode |
149
+ | POST | `/api/afk/stop` | Stop AFK mode |
150
+
151
+ WebSocket events stream execution output in real-time.
152
+
153
+ ## Methodology
154
+
155
+ This tool implements the "Ralph" methodology for AI coding agents:
156
+
157
+ 1. **Small Tasks**: Keep tasks focused and atomic
158
+ 2. **Incremental Progress**: One feature per iteration
159
+ 3. **Progress Tracking**: Maintain a log between sessions
160
+ 4. **Git Commits**: Checkpoint work after each task
161
+ 5. **Feedback Loops**: Use tests and type checking
162
+
163
+ Based on [Anthropic's research on effective harnesses for long-running agents](https://www.anthropic.com/engineering/effective-harnesses-for-long-running-agents).
164
+
165
+ ## License
166
+
167
+ MIT
@@ -0,0 +1 @@
1
+ #!/usr/bin/env node