@su-record/vibe 2.6.15 → 2.6.17

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 CHANGED
@@ -1,18 +1,41 @@
1
1
  # Vibe
2
2
 
3
- **SPEC-Scenario-driven AI coding framework for Claude Code with Multi-LLM orchestration**
3
+ **SPEC-driven AI Coding Framework with Multi-LLM Orchestration**
4
4
 
5
5
  [![npm version](https://img.shields.io/npm/v/@su-record/vibe.svg)](https://www.npmjs.com/package/@su-record/vibe)
6
6
  [![npm downloads](https://img.shields.io/npm/dt/@su-record/vibe)](https://www.npmjs.com/package/@su-record/vibe)
7
7
  [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
8
8
 
9
- ## Install
9
+ > AI coding framework for Claude Code. SPEC-based requirements management, Multi-LLM (Claude + GPT + Gemini) orchestration, and automated quality assurance with 13+ parallel review agents.
10
+
11
+ ## Quick Start
10
12
 
11
13
  ```bash
12
14
  npm install -g @su-record/vibe
13
15
  vibe init
14
16
  ```
15
17
 
18
+ ## Core Workflow
19
+
20
+ ```
21
+ /vibe.spec → /vibe.run → /vibe.trace → /vibe.verify → /vibe.review
22
+ ↓ ↓ ↓ ↓ ↓
23
+ Write SPEC Execute Traceability BDD Verify Parallel Review
24
+ ```
25
+
26
+ ## Key Features
27
+
28
+ | Feature | Description |
29
+ |---------|-------------|
30
+ | **SPEC-driven Development** | Traceable development: Requirements → SPEC → Feature → Test |
31
+ | **Multi-LLM Orchestration** | Claude + GPT + Gemini 3-way verification with automatic routing |
32
+ | **13+ Parallel Review Agents** | Parallel code review: Security, Performance, Architecture, etc. |
33
+ | **ULTRAWORK Mode** | Enable all optimizations with a single keyword |
34
+ | **Fire-and-Forget Agents** | Non-blocking background agent execution |
35
+ | **Phase Pipelining** | Prepare next phase while current phase executes |
36
+ | **Swarm Pattern** | Auto-split complex tasks for parallel processing (v2.7) |
37
+ | **23 Language Presets** | TypeScript, Python, Go, Rust, Swift, Kotlin, and more |
38
+
16
39
  ## Commands
17
40
 
18
41
  ### Terminal
@@ -20,304 +43,75 @@ vibe init
20
43
  | Command | Description |
21
44
  |---------|-------------|
22
45
  | `vibe init` | Initialize project |
23
- | `vibe update` | Update settings |
46
+ | `vibe update` | Update configuration |
24
47
  | `vibe status` | Check status |
25
48
  | `vibe gpt auth` | GPT OAuth authentication |
26
- | `vibe gpt key <KEY>` | GPT API key setup |
27
- | `vibe gpt logout` | GPT logout |
28
49
  | `vibe gemini auth` | Gemini OAuth authentication |
29
- | `vibe gemini key <KEY>` | Gemini API key setup |
30
- | `vibe gemini logout` | Gemini logout |
31
- | `vibe help` | Help |
32
- | `vibe version` | Version info |
33
50
 
34
- ### Claude Code
51
+ ### Claude Code Slash Commands
35
52
 
36
53
  | Command | Description |
37
54
  |---------|-------------|
38
- | `/vibe.spec "feature"` | Create SPEC + parallel research |
55
+ | `/vibe.spec "feature"` | Generate SPEC document + parallel research |
39
56
  | `/vibe.run "feature"` | Execute implementation |
40
57
  | `/vibe.run "feature" ultrawork` | Maximum performance mode |
41
58
  | `/vibe.verify "feature"` | BDD verification |
42
- | `/vibe.trace "feature"` | Requirements traceability matrix (v2.6) |
43
- | `/vibe.review` | 13+ agent parallel review (`--race` for GPT+Gemini) |
44
- | `/vibe.analyze` | Code analysis |
45
- | `/vibe.reason "problem"` | Systematic reasoning |
46
- | `/vibe.utils --ui` | UI preview (Gemini image / ASCII fallback) |
47
- | `/vibe.utils --e2e` | E2E testing, diagrams, etc. |
48
-
49
- ## Workflow
50
-
51
- ```
52
- /vibe.spec → /vibe.run → /vibe.trace → /vibe.verify → /vibe.review
53
-
54
- Coverage check (v2.6)
55
- ```
56
-
57
- ## Key Features
58
-
59
- | Feature | Description |
60
- |---------|-------------|
61
- | **Race Review (v2.6.9)** | GPT + Gemini parallel review with cross-validation |
62
- | **Fire-and-Forget Agents (v2.6)** | Non-blocking background execution with instant handle return |
63
- | **Phase Pipelining (v2.6)** | Prepare next phase during current execution |
64
- | **PRD-to-SPEC Automation (v2.6)** | Auto-generate SPEC from PRD documents |
65
- | **Traceability Matrix (v2.6)** | REQ → SPEC → Feature → Test coverage tracking |
66
- | **Multi-LLM Research (v2.5)** | Claude + GPT + Gemini 3-way validation during SPEC research |
67
- | **Multi-model orchestration** | Claude + GPT-5.2 + Gemini 3 Pro |
68
- | **13+ parallel review agents** | Security, performance, architecture |
69
- | **Iterative SPEC review** | GPT + Gemini convergence loop (max 3 rounds) |
70
- | **Large SPEC auto-split** | 5+ phases auto-split into phase files |
71
- | **BDD auto verification** | Given/When/Then scenario verification |
72
- | **ULTRAWORK mode** | One keyword enables all optimizations |
73
- | **30+ built-in tools** | Code analysis, memory management, quality validation |
74
- | **Auto context management** | 80%+ auto-save, session restore |
75
- | **23 language presets** | TypeScript, Python, Go, Rust, Swift, Kotlin, C#, Ruby, Dart, GDScript |
59
+ | `/vibe.trace "feature"` | Requirements traceability matrix |
60
+ | `/vibe.review` | 13+ agent parallel review |
61
+ | `/vibe.review --race` | GPT + Gemini race review |
76
62
 
77
63
  ## ULTRAWORK Mode
78
64
 
79
- Enable maximum performance with `ultrawork` or `ulw`:
65
+ Activate maximum performance with `ultrawork` or `ulw` keyword:
80
66
 
81
67
  ```bash
82
68
  /vibe.run "feature" ultrawork
83
- /vibe.run "feature" ulw # Same
84
69
  ```
85
70
 
86
- **Enabled features:**
87
- - Parallel subagent exploration (3+ concurrent)
88
- - **Background agents** (v2.6) - Prepare next phase during implementation
89
- - **Phase pipelining** (v2.6) - Zero wait time between phases
90
- - Boulder Loop (auto-progress until all scenarios complete)
91
- - Auto-retry on errors (max 3)
92
- - Continuous execution without phase confirmation
71
+ **Enabled Features:**
72
+ - Parallel sub-agent exploration (3+ concurrent)
73
+ - Background agents (prepare next phase during implementation)
74
+ - Phase pipelining (eliminate wait time between phases)
75
+ - Boulder Loop (auto-continue until all scenarios complete)
93
76
  - Auto-save at 80%+ context
94
77
 
95
- ## Fire-and-Forget Background Manager (v2.6)
96
-
97
- Non-blocking task execution with instant response:
98
-
99
- ```typescript
100
- import { launch, poll, getStats } from '@su-record/vibe/orchestrator';
101
-
102
- // Fire-and-forget: returns immediately (<100ms)
103
- const { taskId } = launch({
104
- prompt: 'Analyze codebase security',
105
- agentName: 'security-agent',
106
- model: 'claude-sonnet-4-5'
107
- });
108
-
109
- // Poll for result later
110
- const result = await poll(taskId);
111
-
112
- // Check queue stats
113
- const stats = getStats();
114
- ```
115
-
116
- **Concurrency limits:**
117
-
118
- | Model | Concurrent Limit |
119
- |--------|------------------|
120
- | Opus | 3 |
121
- | Sonnet | 5 |
122
- | Haiku | 8 |
123
-
124
- **Queue features:**
125
-
126
- - Bounded queue (max 100 tasks)
127
- - Per-model rate limiting
128
- - Task timeout: 180s individual, 10min pipeline
129
- - Automatic cleanup (24h TTL)
130
-
131
- ## Phase Pipelining (v2.6)
132
-
133
- Prepare next phase while current phase executes:
134
-
135
- ```
136
- Phase N execution │ Background: Phase N+1 preparation
137
- ──────────────────────┼──────────────────────────────────
138
- [Implementing...] │ [Analyzing files for next phase]
139
- [Testing...] │ [Pre-generating test cases]
140
- [Complete] │ [Ready to start immediately!]
141
-
142
- Phase N+1 starts with ZERO wait time
143
- ```
144
-
145
- **Speed comparison:**
146
-
147
- | Mode | 5 Phases Total |
148
- |---------------------------|----------------|
149
- | Sequential | ~10min |
150
- | Parallel Exploration | ~7.5min |
151
- | **ULTRAWORK + Pipeline** | **~5min** |
152
-
153
- ## PRD-to-SPEC Automation (v2.6)
154
-
155
- Auto-generate SPEC from PRD documents:
156
-
157
- ```typescript
158
- import { parsePRD } from '@su-record/vibe/tools';
159
-
160
- // Parse PRD (supports Markdown, YAML frontmatter)
161
- const prd = parsePRD(prdContent, 'login-feature');
162
-
163
- // Extracted requirements with auto-generated IDs
164
- // REQ-login-001, REQ-login-002, ...
165
- ```
166
-
167
- **Requirement ID System:**
168
-
169
- - Format: `REQ-{feature}-{number}`
170
- - Auto-deduplication
171
- - Priority inference from keywords (must, should, nice-to-have)
172
-
173
- ## Requirements Traceability Matrix (v2.6)
174
-
175
- Track coverage from requirements to tests:
176
-
177
- ```bash
178
- /vibe.trace "feature-name" # Generate RTM
179
- ```
180
-
181
- **RTM tracks:**
182
-
183
- ```
184
- REQ-login-001 → SPEC Phase 1 → Feature Scenario 1 → login.test.ts
185
- REQ-login-002 → SPEC Phase 2 → Feature Scenario 3 → auth.test.ts
186
- ```
187
-
188
- **Coverage report:**
189
-
190
- - Requirements coverage %
191
- - Missing implementations
192
- - Untested features
193
- - Export: Markdown, HTML
194
-
195
- ## SPEC Versioning (v2.6)
196
-
197
- Automatic version management for SPEC documents:
198
-
199
- ```typescript
200
- import { bumpVersion, generateChangelog } from '@su-record/vibe/tools';
201
-
202
- // Auto-bump version on SPEC changes
203
- await bumpVersion('feature-name', 'minor');
204
-
205
- // Generate changelog
206
- await generateChangelog('feature-name');
207
- ```
208
-
209
- **Features:**
78
+ **Speed Comparison:**
210
79
 
211
- - Semantic versioning (major.minor.patch)
212
- - Git tag integration
213
- - Automatic changelog generation
214
- - Baseline tagging for releases
80
+ | Mode | 5 Phases |
81
+ |------|----------|
82
+ | Sequential | ~10min |
83
+ | Parallel | ~7.5min |
84
+ | **ULTRAWORK + Pipeline** | **~5min** |
215
85
 
216
- ## UI Preview (`--ui`)
86
+ ## Multi-LLM Orchestration
217
87
 
218
- Generate UI previews from text description or design folder:
88
+ ### Automatic Routing
219
89
 
220
- ```bash
221
- /vibe.utils --ui "login form with email/password"
222
- /vibe.utils --ui ./design/dashboard/
223
- ```
224
-
225
- **Features:**
226
-
227
- - **Gemini enabled**: Generates actual UI mockup image
228
- - **Gemini disabled**: ASCII art fallback
229
- - Supports: HTML, PNG, JPG, CSS, JSON, SVG, MD
230
-
231
- ## Multi-model Orchestration
90
+ Automatically routes to the optimal LLM based on prompt keywords:
232
91
 
233
- **Hook-based automatic routing** - Keywords in your prompt trigger the right LLM automatically:
234
-
235
- | Keyword Pattern | Routes to | Use Case |
236
- |-----------------|-----------|----------|
92
+ | Keyword | Routes to | Use Case |
93
+ |---------|-----------|----------|
237
94
  | `architecture`, `design` | GPT | Architecture review |
238
95
  | `UI`, `UX` | Gemini | UI/UX feedback |
239
- | `debugging`, `find bug` | GPT | Bug analysis |
96
+ | `debugging` | GPT | Bug analysis |
240
97
  | `analyze code` | Gemini | Code review |
241
98
 
242
- **Smart Routing features:**
243
- - Automatic LLM selection based on task type
244
- - Exponential backoff retry (3 attempts, 2s → 4s → 8s)
245
- - Auto fallback chain: GPT ↔ Gemini (if primary fails)
246
- - 5-minute availability cache
247
- - Web search via Gemini (Google Search grounding)
248
-
249
- **API usage:**
250
-
251
- ```javascript
252
- import('@su-record/vibe/lib/gpt').then(g => g.ask('question'))
253
- import('@su-record/vibe/lib/gemini').then(g => g.webSearch('search query'))
254
- ```
255
-
256
- ## Multi-LLM Research (v2.5)
257
-
258
- During `/vibe.spec`, research agents use **3 LLM perspectives** for quality assurance:
259
-
260
- ```text
261
- ┌─────────────────────────────────────────────────────────────┐
262
- │ MULTI-LLM PARALLEL RESEARCH │
263
- ├─────────────────────────────────────────────────────────────┤
264
- │ best-practices-agent │
265
- │ ├── Claude: Core patterns, anti-patterns │
266
- │ ├── GPT: Architecture patterns, code conventions │
267
- │ └── Gemini: Latest trends, framework updates │
268
- ├─────────────────────────────────────────────────────────────┤
269
- │ security-advisory-agent │
270
- │ ├── Claude: OWASP Top 10, security patterns │
271
- │ ├── GPT: CVE database, vulnerability details │
272
- │ └── Gemini: Latest security advisories, patches │
273
- └─────────────────────────────────────────────────────────────┘
274
- ```
275
-
276
- > vibe = Quality Assurance Framework
99
+ ### Race Review
277
100
 
278
- ## Race Review (v2.6.9)
279
-
280
- **Multi-LLM competitive review** - Same task runs on GPT + Gemini in parallel, results cross-validated:
101
+ Run GPT + Gemini in parallel for cross-validation:
281
102
 
282
103
  ```bash
283
- /vibe.review --race # All review types
284
- /vibe.review --race security # Security only
285
- ```
286
-
287
- **Cross-validation confidence:**
288
-
289
- | Agreement | Priority | Action |
290
- |-------------------|----------|-------------------------|
291
- | Both agree (100%) | P1 | High confidence finding |
292
- | One model (50%) | P2 | Needs verification |
293
-
294
- **Output:**
295
-
296
- ```text
297
- 🏁 RACE REVIEW
298
- ├─ GPT: [SQL injection, XSS]
299
- └─ Gemini: [SQL injection, CSRF]
300
-
301
- Cross-validation:
302
- - SQL injection (2/2) → 🔴 P1 (100%)
303
- - XSS (1/2) → 🟡 P2 (50%)
304
- - CSRF (1/2) → 🟡 P2 (50%)
104
+ /vibe.review --race
305
105
  ```
306
106
 
307
- **ULTRAWORK default:** Race review is automatically enabled in ULTRAWORK mode.
308
-
309
- **API usage:**
310
-
311
- ```javascript
312
- import('@su-record/vibe/tools').then(t =>
313
- t.raceReview({ reviewType: 'security', code: 'CODE' })
314
- .then(r => console.log(t.formatRaceResult(r)))
315
- )
316
- ```
107
+ | Agreement | Priority | Action |
108
+ |-----------|----------|--------|
109
+ | Both agree (100%) | P1 | High confidence |
110
+ | One model (50%) | P2 | Needs verification |
317
111
 
318
112
  ## Parallel Review Agents
319
113
 
320
- `/vibe.review` runs 13+ agents concurrently:
114
+ Run 13+ agents simultaneously with `/vibe.review`:
321
115
 
322
116
  | Category | Agents |
323
117
  |----------|--------|
@@ -325,143 +119,112 @@ import('@su-record/vibe/tools').then(t =>
325
119
  | Performance | performance-reviewer, complexity-reviewer |
326
120
  | Architecture | architecture-reviewer, simplicity-reviewer |
327
121
  | Language | python, typescript, rails, react reviewers |
328
- | Context | git-history, test-coverage reviewers |
329
122
 
330
- **Priority:**
331
- - 🔴 P1 (Critical): Blocks merge
332
- - 🟡 P2 (Important): Recommended fix
333
- - 🔵 P3 (Nice-to-have): Backlog
123
+ **Priority System:**
124
+ - P1 (Critical): Blocks merge
125
+ - P2 (Important): Fix recommended
126
+ - P3 (Nice-to-have): Backlog
334
127
 
335
- ## Built-in Tools (25)
128
+ ## Swarm Pattern (v2.7)
336
129
 
337
- | Category | Tools |
338
- |----------|-------|
339
- | **Semantic** | `find_symbol`, `find_references`, `analyze_dependency_graph` |
340
- | **Convention** | `analyze_complexity`, `validate_code_quality`, `check_coupling_cohesion`, `suggest_improvements`, `apply_quality_rules` |
341
- | **Memory** | `start_session`, `save_memory`, `auto_save_context`, `recall_memory`, `search_memories`, `search_memories_advanced`, `list_memories`, `update_memory`, `delete_memory`, `link_memories`, `prioritize_memory`, `get_memory_graph`, `create_memory_timeline`, `get_session_context`, `restore_session_context` |
342
- | **UI** | `preview_ui_ascii` |
343
- | **Time** | `get_current_time` |
130
+ Automatically split complex tasks for parallel processing:
344
131
 
345
- ## Project Structure
132
+ ```typescript
133
+ import { swarm, analyzeTaskComplexity } from '@su-record/vibe/orchestrator';
346
134
 
347
- **Global install (`~/.claude/`):**
135
+ // Analyze complexity
136
+ const analysis = analyzeTaskComplexity('Your prompt');
137
+ console.log(analysis.score); // Split if >= 15
348
138
 
349
- ```text
350
- ~/.claude/
351
- ├── commands/ # Slash commands (7)
352
- ├── agents/ # Review/research agents
353
- ├── skills/ # Auto-activated guides (7)
354
- └── settings.json # Hooks settings
139
+ // Execute swarm
140
+ const result = await swarm({
141
+ prompt: 'Complex task...',
142
+ maxDepth: 2, // Maximum split depth
143
+ splitThreshold: 15, // Complexity threshold
144
+ });
355
145
  ```
356
146
 
357
- **Project-specific (`project/.claude/vibe/`):**
358
-
359
- ```text
360
- .claude/vibe/
361
- ├── specs/ # SPEC documents
362
- ├── features/ # BDD scenarios
363
- ├── rules/ # Coding rules (per stack)
364
- ├── solutions/ # Solution archive
365
- ├── config.json # Project settings
366
- └── constitution.md # Project principles
147
+ **How it works:**
148
+ ```
149
+ Prompt → Complexity Analysis → Split Decision
150
+
151
+ ┌─ Low Direct execution
152
+ └─ High Create subtasks → Parallel processing → Merge results
367
153
  ```
368
154
 
369
- ## Cursor IDE Support (v2.6.8)
370
-
371
- VIBE automatically installs Cursor-compatible assets on `vibe init` and `vibe update`:
372
-
373
- **Installed locations:**
374
-
375
- | Asset | Path | Count |
376
- | ----- | ---- | ----- |
377
- | Subagents | `~/.cursor/agents/*.md` | 12 |
378
- | Skills | `~/.cursor/skills/*/SKILL.md` | 7 |
379
- | Rules Template | `~/.cursor/rules-template/*.mdc` | 5 |
380
-
381
- **12 Review Subagents with optimized model mapping:**
382
-
383
- | Agent Type | Model | Purpose |
384
- | ---------- | ----- | ------- |
385
- | security, architecture, data-integrity | `claude-4.5-sonnet-thinking` | Deep reasoning |
386
- | typescript, python, react, rails | `gpt-5.2-codex` | Code understanding |
387
- | performance, complexity, simplicity, test-coverage, git-history | `gemini-3-flash` | Fast pattern checks |
388
-
389
- **7 Skills (VIBE workflow guides):**
155
+ ## Requirements Traceability (v2.6)
390
156
 
391
- - `vibe-spec` - SPEC creation
392
- - `vibe-run` - Implementation execution
393
- - `vibe-review` - Parallel code review
394
- - `vibe-analyze` - Project analysis
395
- - `vibe-verify` - Requirement verification
396
- - `vibe-reason` - Systematic reasoning
397
- - `vibe-ui` - UI preview utilities
157
+ Track from requirements to tests:
398
158
 
399
- **5 Rules Templates:**
159
+ ```bash
160
+ /vibe.trace "feature"
161
+ ```
400
162
 
401
- - `typescript-standards.mdc` - Type safety, complexity limits
402
- - `react-patterns.mdc` - Component patterns
403
- - `code-quality.mdc` - General quality rules
404
- - `security-checklist.mdc` - OWASP checklist
405
- - `python-standards.mdc` - PEP8, type hints
163
+ ```
164
+ REQ-login-001 SPEC Phase 1 → Feature Scenario 1 → login.test.ts
165
+ REQ-login-002 SPEC Phase 2 → Feature Scenario 3 → auth.test.ts
166
+ ```
406
167
 
407
- **Usage in Cursor:**
168
+ ## Project Structure
408
169
 
409
- ```text
410
- # Invoke subagent
411
- "Use security-reviewer to check this code"
170
+ **Global (`~/.claude/`):**
171
+ ```
172
+ ~/.claude/
173
+ ├── commands/ # Slash commands
174
+ ├── agents/ # Review/research agents
175
+ ├── skills/ # Auto-activated guides
176
+ └── settings.json # Hooks
177
+ ```
412
178
 
413
- # Use skill workflow
414
- "vibe spec login feature"
179
+ **Project (`.claude/vibe/`):**
180
+ ```
181
+ .claude/vibe/
182
+ ├── specs/ # SPEC documents
183
+ ├── features/ # BDD scenarios
184
+ ├── config.json # Project settings
185
+ └── constitution.md
415
186
  ```
416
187
 
417
188
  ## Code Quality Standards
418
189
 
419
190
  | Metric | Limit |
420
191
  |--------|-------|
421
- | Function length | 30 lines recommended, 50 allowed |
422
- | Nesting depth | 3 levels max |
423
- | Parameters | 5 max |
424
- | Cyclomatic complexity | 10 max |
425
-
426
- ## Usage Example
427
-
428
- ```
429
- User: /vibe.spec "login feature"
192
+ | Function length | 30 lines (recommended), 50 lines (allowed) |
193
+ | Nesting depth | 3 levels |
194
+ | Parameters | 5 |
195
+ | Cyclomatic complexity | 10 |
430
196
 
431
- Claude: You're building a login feature! I have a few questions.
432
- 1. Auth method? (email/password, OAuth, Passkey)
433
- 2. Tech stack?
434
- ...
197
+ ## Cursor IDE Support
435
198
 
436
- [Requirements confirmed via conversation]
437
- [4 parallel research agents run]
199
+ Assets auto-installed on `vibe init/update`:
438
200
 
439
- SPEC document created
440
- 📄 .claude/vibe/specs/login.md
441
- 📄 .claude/vibe/features/login.feature
201
+ | Asset | Path | Count |
202
+ |-------|------|-------|
203
+ | Subagents | `~/.cursor/agents/` | 12 |
204
+ | Skills | `~/.cursor/skills/` | 7 |
205
+ | Rules | `~/.cursor/rules-template/` | 23 languages |
442
206
 
443
- Next: /vibe.run "login feature" ultrawork
444
- ```
207
+ ## API Usage
445
208
 
446
- ## Context7 Plugin (Recommended)
209
+ ```typescript
210
+ // Background agent
211
+ import { launch, poll } from '@su-record/vibe/orchestrator';
212
+ const { taskId } = launch({ prompt: 'Analyze code', agentName: 'analyzer' });
213
+ const result = await poll(taskId);
447
214
 
448
- Install context7 plugin for up-to-date library documentation:
215
+ // Swarm pattern (v2.7) - Auto-split complex tasks
216
+ import { swarm } from '@su-record/vibe/orchestrator';
217
+ const result = await swarm({
218
+ prompt: 'Implement login with: 1. UI 2. Validation 3. API 4. Tests',
219
+ maxDepth: 2,
220
+ splitThreshold: 15,
221
+ });
449
222
 
450
- ```bash
451
- /plugin install context7
223
+ // LLM direct call
224
+ import { ask } from '@su-record/vibe/lib/gpt';
225
+ import { webSearch } from '@su-record/vibe/lib/gemini';
452
226
  ```
453
227
 
454
- **Benefits:**
455
-
456
- - **Skill integration**: Auto-invokes context7 for library/API questions without prompting
457
- - **Subagent isolation**: Docs queries run in separate context, preventing main context bloat
458
- - **Knowledge cutoff solution**: Access latest docs for long coding sessions
459
-
460
- **Usage:**
461
-
462
- - Automatic: Ask library questions, vibe's skill triggers context7
463
- - Manual: `/context7:docs <library> [query]`
464
-
465
228
  ## Requirements
466
229
 
467
230
  - Node.js 18.0.0+
@@ -471,27 +234,6 @@ Install context7 plugin for up-to-date library documentation:
471
234
 
472
235
  MIT - [GitHub](https://github.com/su-record/vibe)
473
236
 
474
- ## Vibe Setup (AI Coding)
475
-
476
- This project uses [Vibe](https://github.com/su-record/vibe) AI coding framework.
477
-
478
- ### Collaborator Install
479
-
480
- ```bash
481
- # Global install (recommended)
482
- npm install -g @su-record/vibe
483
- vibe update
484
-
485
- # Or use vibe init to setup
486
- vibe init
487
- ```
488
-
489
- ### Usage
490
-
491
- Use slash commands in Claude Code:
492
- - `/vibe.spec "feature"` - Create SPEC document
493
- - `/vibe.run "feature"` - Execute implementation
494
-
495
237
  ---
496
238
 
497
- If this helps your workflow, consider giving it a star!
239
+ If this helps your workflow, consider giving it a star!
@@ -7,9 +7,6 @@ import { TechStack, StackDetails } from '../types.js';
7
7
  * constitution.md 생성 또는 업데이트
8
8
  */
9
9
  export declare function updateConstitution(vibeDir: string, detectedStacks: TechStack[], stackDetails: StackDetails): void;
10
- /**
11
- * CLAUDE.md 업데이트 (vibe 섹션 추가/교체)
12
- */
13
10
  export declare function updateClaudeMd(projectRoot: string, detectedStacks: TechStack[], isUpdate?: boolean): void;
14
11
  /**
15
12
  * 프로젝트 vibe 폴더 설정
@@ -1 +1 @@
1
- {"version":3,"file":"ProjectSetup.d.ts","sourceRoot":"","sources":["../../../src/cli/setup/ProjectSetup.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAMH,OAAO,EAA8B,SAAS,EAAE,YAAY,EAAE,MAAM,aAAa,CAAC;AASlF;;GAEG;AACH,wBAAgB,kBAAkB,CAChC,OAAO,EAAE,MAAM,EACf,cAAc,EAAE,SAAS,EAAE,EAC3B,YAAY,EAAE,YAAY,GACzB,IAAI,CA8CN;AAED;;GAEG;AACH,wBAAgB,cAAc,CAC5B,WAAW,EAAE,MAAM,EACnB,cAAc,EAAE,SAAS,EAAE,EAC3B,QAAQ,UAAQ,GACf,IAAI,CAoEN;AAED;;GAEG;AACH,wBAAgB,WAAW,CAAC,OAAO,EAAE,MAAM,EAAE,cAAc,EAAE,SAAS,EAAE,EAAE,QAAQ,UAAQ,GAAG,IAAI,CAchG;AAED;;GAEG;AACH,wBAAgB,mBAAmB,CAAC,WAAW,EAAE,MAAM,GAAG,IAAI,CAkC7D;AAED;;GAEG;AACH,wBAAgB,eAAe,CAAC,WAAW,EAAE,MAAM,GAAG,IAAI,CA+BzD;AAoCD;;GAEG;AACH,wBAAgB,YAAY,CAC1B,OAAO,EAAE,MAAM,EACf,cAAc,EAAE,SAAS,EAAE,EAC3B,YAAY,EAAE,YAAY,EAC1B,QAAQ,UAAQ,GACf,IAAI,CA4BN;AAQD;;;;;;GAMG;AACH,wBAAgB,kBAAkB,CAAC,WAAW,EAAE,MAAM,GAAG,IAAI,CA6C5D"}
1
+ {"version":3,"file":"ProjectSetup.d.ts","sourceRoot":"","sources":["../../../src/cli/setup/ProjectSetup.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAMH,OAAO,EAA8B,SAAS,EAAE,YAAY,EAAE,MAAM,aAAa,CAAC;AASlF;;GAEG;AACH,wBAAgB,kBAAkB,CAChC,OAAO,EAAE,MAAM,EACf,cAAc,EAAE,SAAS,EAAE,EAC3B,YAAY,EAAE,YAAY,GACzB,IAAI,CA8CN;AASD,wBAAgB,cAAc,CAC5B,WAAW,EAAE,MAAM,EACnB,cAAc,EAAE,SAAS,EAAE,EAC3B,QAAQ,UAAQ,GACf,IAAI,CA0FN;AAED;;GAEG;AACH,wBAAgB,WAAW,CAAC,OAAO,EAAE,MAAM,EAAE,cAAc,EAAE,SAAS,EAAE,EAAE,QAAQ,UAAQ,GAAG,IAAI,CAchG;AAED;;GAEG;AACH,wBAAgB,mBAAmB,CAAC,WAAW,EAAE,MAAM,GAAG,IAAI,CAkC7D;AAED;;GAEG;AACH,wBAAgB,eAAe,CAAC,WAAW,EAAE,MAAM,GAAG,IAAI,CA+BzD;AAoCD;;GAEG;AACH,wBAAgB,YAAY,CAC1B,OAAO,EAAE,MAAM,EACf,cAAc,EAAE,SAAS,EAAE,EAC3B,YAAY,EAAE,YAAY,EAC1B,QAAQ,UAAQ,GACf,IAAI,CA4BN;AAQD;;;;;;GAMG;AACH,wBAAgB,kBAAkB,CAAC,WAAW,EAAE,MAAM,GAAG,IAAI,CA6C5D"}