@yeongjaeyou/claude-code-config 0.11.0 → 0.12.0

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.
@@ -9,16 +9,23 @@ Break down large work items into manageable, independent issues. Follow project
9
9
  - Scan `.claude/agents/` for custom agents (read YAML frontmatter)
10
10
  - Scan `.claude/skills/` for available skills (read SKILL.md)
11
11
  - Check `.mcp.json` for configured MCP servers
12
- 3. Analyze work: Understand core requirements and objectives
13
- 4. Decompose work: Split major tasks into smaller, manageable sub-tasks or issues. **Aim for optimal count over excessive issues (keep it manageable)**
14
- 5. Analyze dependencies: Identify prerequisite tasks
15
- 6. Suggest milestone name: Propose a milestone to group decomposed tasks
16
- 7. Check related PRs (optional): Run `gh pr list --state closed --limit 20` for similar work references (skip if none)
17
- 8. Output decomposed issues: Display issues with proposed milestone name
18
- 9. Ask about GitHub creation: Use AskUserQuestion to let user decide on milestone and issue creation
19
- - Create milestone: `gh api repos/:owner/:repo/milestones -f title="Milestone Name" -f description="Description"`
20
- - Assign issues with `--milestone` option
21
- 10. **Add issues to GitHub Project (optional)**
12
+ - Detect test frameworks (jest.config.*, pytest.ini, vitest.config.*, pyproject.toml, etc.)
13
+ 3. **Check TDD applicability** (if user hasn't specified):
14
+ - Analyze work type: code implementation vs docs/infra/config
15
+ - If test framework detected + code work Ask: "TDD 기반으로 이슈 생성하시겠어요?"
16
+ - If no test framework Inform: "TDD 반영이 필요없습니다. (이유: 테스트 프레임워크 미감지)"
17
+ - If non-code work (docs/infra) → Inform: "TDD 반영이 필요없습니다. (이유: 비코드 작업)"
18
+ - If TDD selected: Add `<!-- TDD: enabled -->` marker to each issue body
19
+ 4. Analyze work: Understand core requirements and objectives
20
+ 5. Decompose work: Split major tasks into smaller, manageable sub-tasks or issues. **Aim for optimal count over excessive issues (keep it manageable)**
21
+ 6. Analyze dependencies: Identify prerequisite tasks
22
+ 7. Suggest milestone name: Propose a milestone to group decomposed tasks
23
+ 8. Check related PRs (optional): Run `gh pr list --state closed --limit 20` for similar work references (skip if none)
24
+ 9. Output decomposed issues: Display issues with proposed milestone name
25
+ 10. Ask about GitHub creation: Use AskUserQuestion to let user decide on milestone and issue creation
26
+ - Create milestone: `gh api repos/:owner/:repo/milestones -f title="Milestone Name" -f description="Description"`
27
+ - Assign issues with `--milestone` option
28
+ 11. **Add issues to GitHub Project (optional)**
22
29
  - Check for existing projects: `gh project list --owner <owner> --format json`
23
30
  - If no project exists: Display "No project found. You can create one with `/gh:init-project`" and skip
24
31
  - If project exists: Ask user via AskUserQuestion whether to add issues
@@ -46,6 +53,8 @@ Examples (vary by project, for reference only):
46
53
  - **Priority**: `priority: high`, `priority: medium`, `priority: low`
47
54
 
48
55
  ### Description
56
+ <!-- TDD: enabled --> ← Add this marker if TDD was selected in Step 3
57
+
49
58
  **Purpose**: [Why this is needed]
50
59
 
51
60
  **Tasks**:
@@ -7,6 +7,7 @@ Act as an expert developer who systematically analyzes and resolves GitHub issue
7
7
 
8
8
  1. **Analyze Issue**:
9
9
  - Run `gh issue view $ISSUE_NUMBER --json title,body,comments,milestone` to get issue title, body, labels, and milestone
10
+ - **Check TDD marker**: Look for `<!-- TDD: enabled -->` in issue body → Set TDD workflow flag
10
11
  - If milestone exists, run `gh issue list --milestone "<milestone-name>" --json number,title,state` to view related issues and understand overall context
11
12
  - Identify requirements precisely
12
13
 
@@ -36,14 +37,28 @@ Act as an expert developer who systematically analyzes and resolves GitHub issue
36
37
  ```
37
38
  - Skip if Status field does not exist
38
39
 
39
- 5. **Analyze Codebase**: Use sub-agents to analyze the codebase in parallel, identifying relevant files and structure needed to resolve the issue.
40
+ 5. **Analyze Codebase (MANDATORY)**: Before writing any code, spawn multiple Explorer agents in parallel to preserve main context:
41
+ - **Structure agent**: Overall architecture and file relationships
42
+ - **Pattern agent**: Similar implementations in codebase
43
+ - **Dependency agent**: Affected modules and consumers
44
+
45
+ ```
46
+ Task tool with subagent_type=Explore (launch 2-3 agents simultaneously)
47
+ ```
40
48
 
41
49
  6. **Plan Resolution**: Based on analysis results, develop a concrete resolution plan and define work steps.
42
50
 
43
51
  7. **Resolve Issue**: Spawn sub-agents to modify code and implement features according to the plan.
52
+ - **If TDD enabled** (marker detected in Step 1):
53
+ 1. 🔴 RED: Write failing tests first based on requirements
54
+ 2. 🟢 GREEN: Implement minimal code to pass tests
55
+ 3. 🔵 REFACTOR: Clean up while keeping tests green
56
+ - **If TDD not enabled**: Implement features directly according to the plan
44
57
  - **Execution verification required**: For Python scripts, executables, or any runnable code, always execute to verify correct behavior. Do not rely solely on file existence or previous results.
45
58
 
46
- 8. **Write Tests**: Spawn independent sub-agents per file to write unit tests in parallel, achieving at least 80% coverage.
59
+ 8. **Write Tests**:
60
+ - **If TDD enabled**: Verify test coverage meets target (tests already written in Step 7), add missing edge cases if needed
61
+ - **If TDD not enabled**: Spawn independent sub-agents per file to write unit tests in parallel, achieving at least 80% coverage
47
62
 
48
63
  9. **Validate**: Run tests, lint checks, and build verification in parallel using independent sub-agents to validate code quality.
49
64
 
@@ -67,6 +67,25 @@ Utilize MCP servers whenever possible:
67
67
 
68
68
  ### Code Exploration Strategy
69
69
 
70
+ #### Pre-Implementation Exploration (MANDATORY)
71
+ Before writing or modifying code, spawn multiple Explorer agents in parallel:
72
+
73
+ | Agent Focus | Purpose |
74
+ |-------------|---------|
75
+ | **Structure** | Overall architecture and file relationships |
76
+ | **Pattern** | Similar implementations in codebase |
77
+ | **Dependency** | Affected modules and consumers |
78
+
79
+ **Benefits:**
80
+ - Preserves main context window for actual implementation
81
+ - Each agent explores independently without blocking
82
+ - Aggregated insights before touching code
83
+
84
+ **Execution:**
85
+ ```
86
+ Task tool with subagent_type=Explore (launch 2-3 agents simultaneously in single message)
87
+ ```
88
+
70
89
  #### Parallel Exploration with Multiple Tools
71
90
  Combine the following tools in parallel for efficient code exploration:
72
91
 
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@yeongjaeyou/claude-code-config",
3
- "version": "0.11.0",
3
+ "version": "0.12.0",
4
4
  "description": "Claude Code CLI custom commands, agents, and skills",
5
5
  "bin": {
6
6
  "claude-code-config": "./bin/cli.js"