@cwim/kanban 1.1.6 → 1.1.8

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.
Files changed (2) hide show
  1. package/README.md +57 -7
  2. package/package.json +1 -1
package/README.md CHANGED
@@ -2,13 +2,9 @@
2
2
 
3
3
  > Your AI's long-term memory. Visualized.
4
4
 
5
- Stop losing context between Claude sessions. CWIM Kanban gives your AI assistant a persistent memory layer with a beautiful dashboard to watch it work.
5
+ **CWIM Kanban gives Claude Code a persistent memory layer.** It remembers what you were working on, recalls context automatically, and shows you everything on a live dashboard.
6
6
 
7
- ## The Problem
8
-
9
- Ever ask Claude to "continue where we left off" and get a blank stare? That's because Claude has no memory between sessions. Every conversation starts fresh, and complex multi-step work gets lost.
10
-
11
- CWIM Kanban fixes this by giving Claude a persistent task memory that survives across sessions.
7
+ No more "what were we doing again?" between sessions.
12
8
 
13
9
  ## How It Works
14
10
 
@@ -43,7 +39,7 @@ kanban
43
39
  # Start the dashboard
44
40
  kanban
45
41
 
46
- # Add to your Claude Code MCP config (~/.claude/config.json)
42
+ # Add to your Claude Code MCP config (~/.claude/claude.json)
47
43
  {
48
44
  "mcpServers": {
49
45
  "kanban": {
@@ -54,6 +50,60 @@ kanban
54
50
  }
55
51
  ```
56
52
 
53
+ ## Making Claude Use It
54
+
55
+ Just installing the MCP isn't enough — Claude needs instructions to use it.
56
+
57
+ Add a `CLAUDE.md` file to your project root:
58
+
59
+ ```markdown
60
+ ## Task Tracking
61
+ Use the cwim-kanban MCP to track all work in this project.
62
+
63
+ ### Workflow
64
+ 1. **Before starting**: Call `task_recall` with what you're about to work on
65
+ 2. **Starting a task**: Create or move to `in-progress`
66
+ 3. **Making progress**: Append notes with discoveries, decisions, or blockers
67
+ 4. **Finishing**: Move to `done` and append a summary note
68
+ 5. **Blocked**: Move to `blocked` with a note explaining why
69
+
70
+ ### Example
71
+ ```
72
+ // Check if we have existing context
73
+ task_recall({ context: "refactoring auth middleware" })
74
+
75
+ // Create or update task
76
+ task_create({
77
+ title: "Refactor auth middleware",
78
+ description: "Extract JWT validation into separate module",
79
+ status: "in-progress",
80
+ tags: ["refactor", "auth"]
81
+ })
82
+
83
+ // Append progress notes
84
+ task_append_note({
85
+ id: "tf-abc123",
86
+ note: "Discovered edge case with refresh tokens"
87
+ })
88
+
89
+ // Mark complete
90
+ task_move({ id: "tf-abc123", status: "done" })
91
+ task_append_note({
92
+ id: "tf-abc123",
93
+ note: "Completed: Extracted JWT validation, all tests passing"
94
+ })
95
+ ```
96
+
97
+ ### Rules
98
+ - Always check for existing tasks before creating new ones
99
+ - Use tags consistently (e.g., "bug", "feature", "refactor", "docs")
100
+ - Append notes liberally - they build context for future sessions
101
+ - Move tasks to "blocked" immediately when stuck, with explanation
102
+ - Keep task titles concise but descriptive
103
+ ```
104
+
105
+ This makes the behavior automatic — no need to ask Claude every session.
106
+
57
107
  ## Memory Features
58
108
 
59
109
  ### Smart Context Recall
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@cwim/kanban",
3
- "version": "1.1.6",
3
+ "version": "1.1.8",
4
4
  "description": "Minimal Kanban task tracking with MCP integration for Claude Code",
5
5
  "type": "module",
6
6
  "main": "dist/index.js",