sumulige-claude 1.1.2 → 1.2.0

package/.claude/workflow/templates/research.md

@@ -0,0 +1,250 @@
+# Phase 1 Research Prompt Template
+
+> **Role**: NotebookLM Knowledge Engine
+> **Goal**: Transform user idea into comprehensive feasibility analysis
+
+---
+
+## System Context
+
+You are the **Phase 1 Research Engine** for a 5-phase AI-assisted development workflow. Your role is to:
+
+1. **Understand** the user's idea/requirement deeply
+2. **Connect dots** by finding related work, patterns, and decisions
+3. **Research** industry best practices using both local knowledge and web search
+4. **Assess** feasibility with concrete technical analysis
+
+**Key Principle**: You provide **zero-hallucination** analysis. If you don't know something from the knowledge base or web search, explicitly state it.
+
+---
+
+## Input Format
+
+```
+USER IDEA: {{userIdea}}
+
+CONTEXT:
+{{additionalContext}}
+```
+
+---
+
+## Research Process
+
+### Step 1: Requirement Structuring (30 minutes)
+
+Extract and clarify:
+- **Core problem** being solved
+- **Target users** and their pain points
+- **Key features** requested
+- **Constraints** (technical, time, resources)
+- **Assumptions** made
+
+Output format:
+```markdown
+## Requirements Summary
+
+### Problem Statement
+[What problem are we solving?]
+
+### Target Users
+[Who will use this? What are their pain points?]
+
+### Key Features
+1. [Feature 1]
+2. [Feature 2]
+...
+
+### Constraints
+- [Constraint 1]
+- [Constraint 2]
+
+### Assumptions
+- [Assumption 1]
+- [Assumption 2]
+```
+
+---
+
+### Step 2: Correlation Analysis (Connect The Dots) (45 minutes)
+
+Search the knowledge base for:
+- **Related projects** with similar features
+- **Reusable components** and patterns
+- **Historical decisions** and their outcomes
+- **Lessons learned** from past projects
+
+Output format:
+```markdown
+## Correlation Analysis
+
+### Research Plan (研究计划)
+- **Key Objectives**: 每个选项要回答的核心问题，评估需要的数据/信息
+- **Research Methods**: 收集和分析数据的方法，使用的工具或方法论
+- **Evaluation Criteria**: 比较选项的指标/基准，可行性判断标准
+- **Expected Outcomes**: 预期的研究发现，研究后的下一步行动
+
+### Related Projects
+| Project | Similarity | Reusable Components |
+|---------|------------|---------------------|
+| [Project A] | 85% | [Component list] |
+| [Project B] | 60% | [Component list] |
+
+### Overlapping Technology
+- [Tech stack overlap]
+- [Shared libraries]
+- [Common patterns]
+
+### Lessons from History
+- [Lesson 1]: [Context]
+- [Lesson 2]: [Context]
+
+### Risks Based on History
+- [Risk from past project]: [How we'll address it]
+```
+
+**Completion Criteria (完成验收标准)**:
+
+完成关联分析后，应能够回答以下问题：
+
+1. **Key Objectives (关键目标)**
+   - 每个选项要回答的核心问题是什么？
+   - 评估需要哪些数据/信息？
+
+2. **Research Methods (研究方法)**
+   - 如何收集和分析数据？
+   - 使用的工具或方法论是什么？
+
+3. **Evaluation Criteria (评估标准)**
+   - 比较选项的指标、基准或定性因素是什么？
+   - 可行性/成功的判断标准是什么？
+
+4. **Expected Outcomes (预期成果)**
+   - 可能的研究发现或结果是什么？
+   - 研究后的下一步行动是什么？
+
+如果以上问题都有明确答案，则 Step 2 完成。
+
+---
+
+### Step 3: Best Practices Research (45 minutes)
+
+For each key area (architecture, tech stack, security, UX, etc.):
+
+1. **Query local knowledge base** first
+2. **Search web** for latest practices if needed
+3. **Synthesize** recommendations
+
+Output format:
+```markdown
+## Industry Best Practices
+
+### [Area: e.g., Frontend Architecture]
+**Practice**: [Specific practice]
+**Rationale**: [Why this is recommended]
+**Sources**: [Citations]
+**Applicability**: [How this applies to current project]
+
+### [Area: e.g., API Design]
+...
+```
+
+---
+
+### Step 4: Feasibility Assessment (30 minutes)
+
+Assess across multiple dimensions:
+
+```markdown
+## Feasibility Assessment
+
+### Technical Feasibility: ⭐⭐⭐⭐☆ (4/5)
+**Strengths**:
+- [Strength 1]
+- [Strength 2]
+
+**Challenges**:
+- [Challenge 1]: [Mitigation strategy]
+- [Challenge 2]: [Mitigation strategy]
+
+### Time Estimate: 4 hours
+**Breakdown**:
+- Research & Planning: 30m
+- Design: 1h
+- Implementation: 2h
+- Testing: 30m
+
+### Complexity: Medium
+**Reasoning**: [Explain complexity assessment]
+
+### Risk Assessment
+| Risk | Severity | Mitigation |
+|------|----------|------------|
+| [Risk 1] | High | [Mitigation] |
+| [Risk 2] | Medium | [Mitigation] |
+```
+
+---
+
+### Step 5: Recommendations (15 minutes)
+
+```markdown
+## Recommendations
+
+### Recommended Tech Stack
+- [Frontend]: [Choice + rationale]
+- [Backend]: [Choice + rationale]
+- [Database]: [Choice + rationale]
+- [Other]: [Choice + rationale]
+
+### Suggested Architecture
+[High-level architecture description]
+
+### Potential Issues to Watch
+1. [Issue 1]: [Monitoring approach]
+2. [Issue 2]: [Monitoring approach]
+
+### Next Steps (Phase 2)
+1. [Step 1]
+2. [Step 2]
+3. [Step 3]
+```
+
+---
+
+## Quality Checklist
+
+Before finalizing, ensure:
+
+- [ ] Requirement summary is clear and complete
+- [ ] Correlation analysis found related work/patterns
+- [ ] Best practices are cited with sources
+- [ ] Feasibility has concrete ratings (not vague)
+- [ ] Time estimate is justified
+- [ ] Risks have mitigation strategies
+- [ ] Recommendations are actionable
+
+---
+
+## Output Format
+
+Final output should be saved as `feasibility-report.md` in:
+```
+development/projects/{projectId}/phase1/feasibility-report.md
+```
+
+---
+
+## Example Output
+
+See `development/projects/examples/phase1/feasibility-report.md` for a complete example.
+
+---
+
+## Notes for AI
+
+- **Be specific**: Use concrete examples, not abstract advice
+- **Cite sources**: Always reference where information comes from
+- **Quantify**: Use numbers when possible (similarity scores, time estimates)
+- **Be honest**: If knowledge is missing, state it explicitly
+- **Think ahead**: Consider Phase 2 (approval) - what will Claude need to know?

package/.claude/workflow/types.js

@@ -0,0 +1,37 @@
+/**
+ * Workflow Type Definitions (JavaScript version)
+ * Shared types for the NotebookLM + Claude collaboration workflow
+ */
+
+// Project phase enum
+const ProjectPhase = {
+  RESEARCH: 'research',      // Phase 1: NotebookLM feasibility analysis
+  APPROVAL: 'approval',      // Phase 2: Claude review and consensus
+  PLANNING: 'planning',      // Phase 3: PRD and prototype
+  DEVELOPMENT: 'development',// Phase 4: Implementation
+  DEPLOYMENT: 'deployment'   // Phase 5: Release
+};
+
+// Project status
+const ProjectStatus = {
+  PENDING: 'pending',
+  IN_PROGRESS: 'in_progress',
+  COMPLETED: 'completed',
+  FAILED: 'failed',
+  ON_HOLD: 'on_hold'
+};
+
+// Knowledge source types
+const KnowledgeSourceType = {
+  LOCAL_FILE: 'local_file',
+  LOCAL_DIRECTORY: 'local_directory',
+  NOTEBOOKLM: 'notebooklm',
+  WEB_SEARCH: 'web_search',
+  WEB_URL: 'web_url'
+};
+
+module.exports = {
+  ProjectPhase,
+  ProjectStatus,
+  KnowledgeSourceType
+};

package/.claude/workflow/web-search.js

@@ -0,0 +1,278 @@
+/**
+ * Web Search Module - Bing HTML-based search
+ *
+ * Provides free web search by parsing Bing's HTML results.
+ */
+
+const https = require('https');
+
+// ============================================================================
+// Configuration
+// ============================================================================
+
+const SEARCH_TIMEOUT = 10000;
+const MAX_RESULTS = 5;
+const USER_AGENT = 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36';
+
+// ============================================================================
+// Web Search Class
+// ============================================================================
+
+class WebSearch {
+  /**
+   * Search Bing and return results
+   */
+  static async search(query, options = {}) {
+    const {
+      maxResults = MAX_RESULTS,
+      timeout = SEARCH_TIMEOUT
+    } = options;
+
+    return new Promise((resolve, reject) => {
+      const encodedQuery = encodeURIComponent(query);
+      const searchUrl = `https://www.bing.com/search?q=${encodedQuery}&setlang=en`;
+
+      const urlObj = new URL(searchUrl);
+      const requestOptions = {
+        hostname: urlObj.hostname,
+        path: urlObj.pathname + urlObj.search,
+        method: 'GET',
+        headers: {
+          'User-Agent': USER_AGENT,
+          'Accept': 'text/html',
+          'Accept-Language': 'en-US,en;q=0.9'
+        },
+        timeout: timeout
+      };
+
+      const req = https.get(requestOptions, (res) => {
+        // Handle redirects
+        if (res.statusCode === 301 || res.statusCode === 302) {
+          const redirectUrl = res.headers.location;
+          if (redirectUrl) {
+            this.followRedirect(redirectUrl, options)
+              .then(resolve)
+              .catch(reject);
+            return;
+          }
+        }
+
+        if (res.statusCode !== 200) {
+          reject(new Error(`HTTP ${res.statusCode}`));
+          return;
+        }
+
+        let data = '';
+        res.on('data', (chunk) => {
+          data += chunk;
+        });
+
+        res.on('end', () => {
+          try {
+            const results = this.parseHTML(data, maxResults);
+            resolve(results);
+          } catch (error) {
+            reject(new Error(`Parse error: ${error.message}`));
+          }
+        });
+      });
+
+      req.on('timeout', () => {
+        req.destroy();
+        reject(new Error('Request timed out'));
+      });
+
+      req.on('error', (error) => {
+        reject(new Error(`Request failed: ${error.message}`));
+      });
+    });
+  }
+
+  /**
+   * Follow HTTP redirect
+   */
+  static followRedirect(url, options) {
+    return new Promise((resolve, reject) => {
+      const urlObj = new URL(url);
+      const isHttps = urlObj.protocol === 'https:';
+      const http = isHttps ? https : require('http');
+
+      const requestOptions = {
+        hostname: urlObj.hostname,
+        path: urlObj.pathname + urlObj.search,
+        method: 'GET',
+        headers: { 'User-Agent': USER_AGENT },
+        timeout: options.timeout || SEARCH_TIMEOUT
+      };
+
+      const req = http.get(requestOptions, (res) => {
+        if (res.statusCode === 301 || res.statusCode === 302) {
+          const redirectUrl = res.headers.location;
+          if (redirectUrl) {
+            this.followRedirect(redirectUrl, options)
+              .then(resolve)
+              .catch(reject);
+            return;
+          }
+        }
+
+        let data = '';
+        res.on('data', (chunk) => { data += chunk; });
+        res.on('end', () => {
+          try {
+            const results = this.parseHTML(data, options.maxResults || MAX_RESULTS);
+            resolve(results);
+          } catch (error) {
+            reject(error);
+          }
+        });
+      });
+
+      req.on('error', reject);
+    });
+  }
+
+  /**
+   * Parse Bing HTML results
+   */
+  static parseHTML(html, maxResults) {
+    const results = [];
+
+    // Bing result patterns
+    // Look for <li class="b_algo"> or <li class="b_result">
+    const algoRegex = /<li[^>]*class="b_algo"[^>]*>([\s\S]*?)<\/li>/g;
+    let match;
+    let count = 0;
+
+    while ((match = algoRegex.exec(html)) !== null && count < maxResults) {
+      const block = match[1];
+      count++;
+
+      // Extract title (from <h2><a> tag)
+      const titleMatch = block.match(/<h2><a[^>]*>([^<]+)<\/a>/);
+      let title = titleMatch ? this.decodeHTML(titleMatch[1]) : '';
+
+      // Also try alternative pattern
+      if (!title) {
+        const altTitleMatch = block.match(/<a[^>]*h="[^"]*"[^>]*>([^<]*(?:<strong>[^<]*<\/strong>[^<]*)*)<\/a>/);
+        if (altTitleMatch) {
+          title = this.decodeHTML(altTitleMatch[1].replace(/<strong>/g, '').replace(/<\/strong>/g, ''));
+        }
+      }
+
+      // Extract URL
+      let url = '';
+      const urlMatch = block.match(/<a[^>]*href="([^"]+)"/);
+      if (urlMatch) {
+        url = urlMatch[1];
+
+        // Decode &amp; entities in URL
+        url = url.replace(/&amp;/g, '&');
+
+        // Clean up Bing redirect URLs
+        // Format: https://www.bing.com/ck/a?!&p=...&u=<encoded_url>
+        if (url.includes('bing.com/ck/a')) {
+          const uMatch = url.match(/[?&]u=([^&]+)/);
+          if (uMatch) {
+            let encodedUrl = uMatch[1];
+            try {
+              // The URL is double-encoded, decode twice
+              let decodedUrl = decodeURIComponent(encodedUrl);
+              decodedUrl = decodeURIComponent(decodedUrl);
+
+              // Bing sometimes adds 'a1' prefix to the final URL
+              // Remove it and decode the rest as base64
+              if (decodedUrl.startsWith('a1')) {
+                const base64Part = decodedUrl.substring(2);
+                decodedUrl = Buffer.from(base64Part, 'base64').toString('utf-8');
+              }
+
+              url = decodedUrl;
+            } catch (e) {
+              // If decode fails, try alternative
+              try {
+                if (encodedUrl.startsWith('a1')) {
+                  const base64Part = encodedUrl.substring(2);
+                  url = Buffer.from(base64Part, 'base64').toString('utf-8');
+                } else {
+                  url = decodeURIComponent(encodedUrl);
+                }
+              } catch (e2) {}
+            }
+          }
+        }
+      }
+
+      // Extract snippet (from <p> or <div>)
+      let snippet = '';
+      const snippetMatch = block.match(/<p[^>]*>([\s\S]*?)<\/p>/);
+      if (snippetMatch) {
+        snippet = this.decodeHTML(snippetMatch[1]
+          .replace(/<strong>/g, '').replace(/<\/strong>/g, '')
+          .replace(/<b>/g, '').replace(/<\/b>/g, '')
+          .replace(/<[^>]+>/g, ' ')
+          .replace(/\s+/g, ' ')
+          .trim());
+      }
+
+      // Skip if no valid URL
+      if (!url || url.startsWith('#')) continue;
+
+      results.push({
+        title: title || 'Untitled',
+        url,
+        excerpt: snippet || 'No description available.',
+        source: 'bing'
+      });
+    }
+
+    return results;
+  }
+
+  /**
+   * Decode HTML entities
+   */
+  static decodeHTML(text) {
+    const entities = {
+      '&amp;': '&',
+      '&lt;': '<',
+      '&gt;': '>',
+      '&quot;': '"',
+      '&#39;': "'",
+      '&apos;': "'",
+      '&nbsp;': ' ',
+      '&middot;': '·'
+    };
+
+    return text.replace(/&[^;]+;/g, (entity) => {
+      return entities[entity] || entity;
+    });
+  }
+
+  /**
+   * Deduplicate results by URL
+   */
+  static dedupeResults(results) {
+    const seen = new Set();
+    return results.filter(result => {
+      const normalizedUrl = result.url
+        .replace(/^https?:\/\//, '')
+        .replace(/\/$/, '')
+        .toLowerCase();
+
+      if (seen.has(normalizedUrl)) return false;
+      seen.add(normalizedUrl);
+      return true;
+    });
+  }
+}
+
+// ============================================================================
+// Exports
+// ============================================================================
+
+module.exports = {
+  WebSearch,
+  SEARCH_TIMEOUT,
+  MAX_RESULTS
+};

package/.claude-plugin/marketplace.json

@@ -27,8 +27,8 @@
     }
   ],
   "metadata": {
-    "version": "1.0.11",
-    "generated_at": "2026-01-15T08:24:34.719Z",
+    "version": "1.1.2",
+    "generated_at": "2026-01-17T09:45:22.671Z",
     "skill_count": 1,
     "categories": {
       "tools": {