claude-queue 1.1.0 → 1.2.1
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/dist/skills/queue/SKILL.md +15 -29
- package/package.json +1 -1
|
@@ -41,35 +41,21 @@ Check the arguments passed to this skill:
|
|
|
41
41
|
|
|
42
42
|
When invoked with `/queue plan <project-id>`:
|
|
43
43
|
|
|
44
|
-
1. **Connect**: Call `queue_watch` with the project ID
|
|
45
|
-
2. **
|
|
46
|
-
3. **
|
|
47
|
-
|
|
48
|
-
-
|
|
49
|
-
-
|
|
50
|
-
|
|
51
|
-
|
|
52
|
-
|
|
53
|
-
|
|
54
|
-
|
|
55
|
-
|
|
56
|
-
|
|
57
|
-
|
|
58
|
-
|
|
59
|
-
Claude: What would you like to plan?
|
|
60
|
-
User: User authentication with email/password
|
|
61
|
-
Claude: Here's my proposed breakdown:
|
|
62
|
-
|
|
63
|
-
1. **Create auth database schema** - users table with email, password_hash, created_at
|
|
64
|
-
2. **Implement registration endpoint** - POST /auth/register with validation
|
|
65
|
-
3. **Implement login endpoint** - POST /auth/login returning JWT
|
|
66
|
-
4. **Add auth middleware** - Verify JWT on protected routes
|
|
67
|
-
5. **Create React auth context** - Login state and token management
|
|
68
|
-
|
|
69
|
-
Want me to adjust anything? If not, should I add these to ready (default) or backlog?
|
|
70
|
-
User: Looks good, add to ready
|
|
71
|
-
Claude: ✓ Created 5 tasks in ready column for kbn-a3x9
|
|
72
|
-
```
|
|
44
|
+
1. **Connect**: Call `queue_watch` with the project ID (ignore the "watching" message - that's for work mode)
|
|
45
|
+
2. **Start discovery**: Immediately ask the user an open-ended question about what they're planning to work on. Let them explain freely.
|
|
46
|
+
3. **Gather context**: After the user replies, read the project's CLAUDE.md file if it exists (at the project path) to understand any existing context
|
|
47
|
+
4. **Iterative refinement**: Use the AskUserQuestion tool to ask targeted follow-up questions based on their initial response and any existing context. Keep asking until you have a clear understanding of:
|
|
48
|
+
- What they want to accomplish
|
|
49
|
+
- The scope and constraints
|
|
50
|
+
- What success looks like
|
|
51
|
+
- Any preferences or requirements
|
|
52
|
+
5. **Propose the plan**: Present a comprehensive plan document capturing everything discussed. Ask if it captures their intent correctly.
|
|
53
|
+
6. **Refine until approved**: If the user wants changes, refine the plan. Repeat until they're happy.
|
|
54
|
+
7. **Create/update CLAUDE.md**: Once approved, write the plan to CLAUDE.md in the project directory. This provides context for all future sessions.
|
|
55
|
+
8. **Break into tasks**: Derive task titles and detailed descriptions from the plan
|
|
56
|
+
9. **Ready to start?**: Ask "Ready to start? (or should I just add these to backlog for later?)"
|
|
57
|
+
- If user wants to start: Create tasks in "ready", then proceed to **Work Mode** and begin the main loop
|
|
58
|
+
- If user wants backlog: Create tasks in "backlog", inform them they can start later by running `/queue <project-id>`
|
|
73
59
|
|
|
74
60
|
---
|
|
75
61
|
|