autoworkflow 3.1.4 → 3.5.0

package/.claude/skills/mcp-github.md

@@ -0,0 +1,217 @@
+# MCP GitHub Skill
+
+## Server Configuration
+\`\`\`json
+// claude_desktop_config.json or .claude/settings.json
+{
+  "mcpServers": {
+    "github": {
+      "command": "npx",
+      "args": ["-y", "@modelcontextprotocol/server-github"],
+      "env": {
+        "GITHUB_PERSONAL_ACCESS_TOKEN": "ghp_xxxxxxxxxxxx"
+      }
+    }
+  }
+}
+\`\`\`
+
+## Authentication Setup
+\`\`\`bash
+# Create a GitHub Personal Access Token (PAT)
+# Settings > Developer settings > Personal access tokens > Fine-grained tokens
+
+# Required scopes for full functionality:
+# - Repository access: All repositories (or specific repos)
+# - Permissions:
+#   - Contents: Read and write
+#   - Issues: Read and write
+#   - Pull requests: Read and write
+#   - Metadata: Read-only
+
+# Store token securely
+export GITHUB_PERSONAL_ACCESS_TOKEN="ghp_xxxxxxxxxxxx"
+\`\`\`
+
+## Available Tools
+
+### Repository Operations
+\`\`\`
+get_file_contents
+- owner: string (repo owner)
+- repo: string (repository name)
+- path: string (file path)
+- branch?: string (optional branch)
+
+push_files
+- owner: string
+- repo: string
+- branch: string
+- files: array of { path, content }
+- message: string (commit message)
+
+create_repository
+- name: string
+- description?: string
+- private?: boolean
+- auto_init?: boolean
+
+fork_repository
+- owner: string
+- repo: string
+
+create_branch
+- owner: string
+- repo: string
+- branch: string
+- from_branch?: string
+
+search_repositories
+- query: string (GitHub search syntax)
+- page?: number
+- per_page?: number
+
+search_code
+- query: string
+- page?: number
+- per_page?: number
+\`\`\`
+
+### Issue Operations
+\`\`\`
+create_issue
+- owner: string
+- repo: string
+- title: string
+- body?: string
+- assignees?: string[]
+- labels?: string[]
+- milestone?: number
+
+update_issue
+- owner: string
+- repo: string
+- issue_number: number
+- title?: string
+- body?: string
+- state?: 'open' | 'closed'
+- assignees?: string[]
+- labels?: string[]
+
+get_issue
+- owner: string
+- repo: string
+- issue_number: number
+
+list_issues
+- owner: string
+- repo: string
+- state?: 'open' | 'closed' | 'all'
+- labels?: string
+- sort?: 'created' | 'updated' | 'comments'
+- direction?: 'asc' | 'desc'
+
+add_issue_comment
+- owner: string
+- repo: string
+- issue_number: number
+- body: string
+\`\`\`
+
+### Pull Request Operations
+\`\`\`
+create_pull_request
+- owner: string
+- repo: string
+- title: string
+- body?: string
+- head: string (source branch)
+- base: string (target branch)
+- draft?: boolean
+
+get_pull_request
+- owner: string
+- repo: string
+- pull_number: number
+
+list_pull_requests
+- owner: string
+- repo: string
+- state?: 'open' | 'closed' | 'all'
+- head?: string
+- base?: string
+- sort?: 'created' | 'updated' | 'popularity'
+
+create_review
+- owner: string
+- repo: string
+- pull_number: number
+- body?: string
+- event: 'APPROVE' | 'REQUEST_CHANGES' | 'COMMENT'
+- comments?: array (inline comments)
+
+merge_pull_request
+- owner: string
+- repo: string
+- pull_number: number
+- commit_title?: string
+- merge_method?: 'merge' | 'squash' | 'rebase'
+\`\`\`
+
+## Common Workflows
+
+### Create Issue from Bug Report
+\`\`\`
+1. Use create_issue with detailed description
+2. Add appropriate labels (bug, priority)
+3. Assign to team member
+4. Link to related issues in body
+\`\`\`
+
+### Review and Merge PR
+\`\`\`
+1. Use get_pull_request to fetch PR details
+2. Use get_file_contents to review changed files
+3. Use create_review to approve or request changes
+4. Use merge_pull_request with squash method
+\`\`\`
+
+### Search and Analyze Code
+\`\`\`
+1. Use search_code with GitHub search syntax
+   - "filename:*.ts extension:ts"
+   - "function useAuth language:typescript"
+2. Use get_file_contents to read matching files
+3. Analyze patterns across repositories
+\`\`\`
+
+## GitHub Search Syntax
+\`\`\`
+# Repository search
+stars:>1000 language:typescript
+topic:react pushed:>2024-01-01
+
+# Code search
+filename:package.json "react"
+extension:ts "interface User"
+repo:owner/name path:src/ "function"
+
+# Issue search
+is:issue is:open label:bug
+author:username created:>2024-01-01
+\`\`\`
+
+## ❌ DON'T
+- Commit PAT tokens to repositories
+- Use overly broad token permissions
+- Create spam issues or PRs
+- Push directly to protected branches
+- Forget to handle rate limiting
+
+## ✅ DO
+- Use fine-grained PATs with minimal scopes
+- Store tokens in environment variables
+- Use meaningful commit messages
+- Follow repository contributing guidelines
+- Handle pagination for large result sets
+- Respect rate limits (5000 requests/hour)

package/.claude/skills/mcp-memory.md

@@ -0,0 +1,240 @@
+# MCP Memory Skill
+
+## Server Configuration
+\`\`\`json
+// claude_desktop_config.json or .claude/settings.json
+{
+  "mcpServers": {
+    "memory": {
+      "command": "npx",
+      "args": ["-y", "@modelcontextprotocol/server-memory"]
+    }
+  }
+}
+\`\`\`
+
+## Knowledge Graph Architecture
+The memory server uses a knowledge graph model with:
+- **Entities**: Nodes representing concepts, people, projects, etc.
+- **Relations**: Edges connecting entities with named relationships
+- **Observations**: Facts or attributes about entities
+
+## Available Tools
+
+### Entity Operations
+\`\`\`
+create_entities
+- entities: array of {
+    name: string (unique identifier),
+    entityType: string (e.g., "person", "project", "concept"),
+    observations: string[] (facts about the entity)
+  }
+- Returns: created entities
+
+delete_entities
+- entityNames: string[]
+- Returns: success status
+
+read_graph
+- Returns: full knowledge graph with all entities and relations
+
+search_nodes
+- query: string (search term)
+- Returns: matching entities and their relations
+
+open_nodes
+- names: string[]
+- Returns: specified entities with all observations and relations
+\`\`\`
+
+### Relation Operations
+\`\`\`
+create_relations
+- relations: array of {
+    from: string (source entity name),
+    to: string (target entity name),
+    relationType: string (e.g., "works_on", "knows", "part_of")
+  }
+- Returns: created relations
+
+delete_relations
+- relations: array of { from, to, relationType }
+- Returns: success status
+\`\`\`
+
+### Observation Operations
+\`\`\`
+add_observations
+- observations: array of {
+    entityName: string,
+    contents: string[] (new facts to add)
+  }
+- Returns: updated entities
+
+delete_observations
+- deletions: array of {
+    entityName: string,
+    observations: string[] (facts to remove)
+  }
+- Returns: success status
+\`\`\`
+
+## Knowledge Patterns
+
+### Person Profile
+\`\`\`
+Entity: "John Smith"
+Type: "person"
+Observations:
+  - "Senior software engineer at TechCorp"
+  - "Prefers TypeScript over JavaScript"
+  - "Works remotely from Seattle"
+  - "Expert in React and Node.js"
+
+Relations:
+  - John Smith --works_on--> "Project Alpha"
+  - John Smith --manages--> "Frontend Team"
+  - John Smith --knows--> "Jane Doe"
+\`\`\`
+
+### Project Knowledge
+\`\`\`
+Entity: "Project Alpha"
+Type: "project"
+Observations:
+  - "E-commerce platform rebuild"
+  - "Stack: Next.js, PostgreSQL, Redis"
+  - "Launch target: Q2 2025"
+  - "Budget: $500K"
+
+Relations:
+  - Project Alpha --uses--> "Next.js"
+  - Project Alpha --deployed_on--> "AWS"
+  - Project Alpha --owned_by--> "Engineering Team"
+\`\`\`
+
+### Technical Preferences
+\`\`\`
+Entity: "User Preferences"
+Type: "preferences"
+Observations:
+  - "Prefers functional components over class components"
+  - "Uses Prettier with single quotes"
+  - "Tab size: 2 spaces"
+  - "Testing framework: Vitest"
+  - "State management: Zustand"
+\`\`\`
+
+### Codebase Context
+\`\`\`
+Entity: "MyApp Codebase"
+Type: "codebase"
+Observations:
+  - "Monorepo using pnpm workspaces"
+  - "Authentication via Clerk"
+  - "API routes in /app/api"
+  - "Components in /src/components"
+  - "Shared types in /packages/types"
+
+Relations:
+  - MyApp Codebase --depends_on--> "PostgreSQL DB"
+  - MyApp Codebase --deployed_to--> "Vercel"
+\`\`\`
+
+## Common Workflows
+
+### Initial Setup
+\`\`\`
+1. Create core entities:
+   - User profile (preferences, expertise)
+   - Project (name, stack, goals)
+   - Codebase (structure, patterns)
+
+2. Add relations:
+   - User --works_on--> Project
+   - Project --uses--> Technologies
+
+3. Record observations as you learn them
+\`\`\`
+
+### During Conversation
+\`\`\`
+1. search_nodes for relevant context
+2. open_nodes to get full details
+3. add_observations for new information
+4. create_relations for new connections
+\`\`\`
+
+### Context Retrieval
+\`\`\`
+1. read_graph for overview
+2. search_nodes with specific query
+3. Use context in responses
+\`\`\`
+
+## Entity Types Reference
+\`\`\`
+person      - Team members, stakeholders
+project     - Software projects, initiatives
+codebase    - Repository structure, patterns
+technology  - Frameworks, libraries, tools
+concept     - Technical concepts, patterns
+preferences - User preferences, settings
+decision    - Architectural decisions, rationale
+task        - Ongoing work, todos
+issue       - Known bugs, problems
+\`\`\`
+
+## Relation Types Reference
+\`\`\`
+works_on    - Person works on project
+manages     - Person manages team/project
+knows       - Person knows person
+uses        - Project/codebase uses technology
+depends_on  - Component depends on component
+part_of     - Entity is part of larger entity
+created_by  - Entity created by person
+deployed_to - Project deployed to platform
+replaces    - New tech replaces old tech
+related_to  - General relation
+\`\`\`
+
+## Best Practices for Memory
+
+### Naming Conventions
+\`\`\`
+# Use consistent, descriptive names
+Good: "React Router v6", "Auth Service", "John Smith"
+Bad: "router", "auth", "john"
+
+# Include version/context when relevant
+Good: "PostgreSQL 15 Config"
+Bad: "DB Config"
+\`\`\`
+
+### Observation Quality
+\`\`\`
+# Be specific and actionable
+Good: "Uses ESLint with Airbnb config"
+Bad: "Has linting"
+
+# Include reasoning when relevant
+Good: "Chose Zustand over Redux for simpler API"
+Bad: "Uses Zustand"
+\`\`\`
+
+## ❌ DON'T
+- Store sensitive data (passwords, tokens)
+- Create duplicate entities with different names
+- Add vague, unhelpful observations
+- Create circular or conflicting relations
+- Forget to update outdated information
+
+## ✅ DO
+- Use consistent entity naming
+- Add context-rich observations
+- Create meaningful relations
+- Update entities as information changes
+- Search before creating new entities
+- Delete outdated observations
+- Use appropriate entity types

package/.claude/skills/mcp-search.md

@@ -0,0 +1,218 @@
+# MCP Search Skill
+
+## Server Configuration (Brave Search)
+\`\`\`json
+// claude_desktop_config.json or .claude/settings.json
+{
+  "mcpServers": {
+    "brave-search": {
+      "command": "npx",
+      "args": ["-y", "@modelcontextprotocol/server-brave-search"],
+      "env": {
+        "BRAVE_API_KEY": "BSA..."
+      }
+    }
+  }
+}
+\`\`\`
+
+## Server Configuration (Tavily)
+\`\`\`json
+{
+  "mcpServers": {
+    "tavily": {
+      "command": "npx",
+      "args": ["-y", "@tavily/mcp-server"],
+      "env": {
+        "TAVILY_API_KEY": "tvly-..."
+      }
+    }
+  }
+}
+\`\`\`
+
+## Available Tools
+
+### Web Search
+\`\`\`
+brave_web_search
+- query: string (search query)
+- count?: number (results count, default 10, max 20)
+- offset?: number (pagination offset)
+- Returns: array of {
+    title: string,
+    url: string,
+    description: string,
+    age?: string (how old the result is)
+  }
+
+tavily_search
+- query: string
+- search_depth?: 'basic' | 'advanced'
+- include_domains?: string[]
+- exclude_domains?: string[]
+- max_results?: number
+- Returns: array of search results with content snippets
+\`\`\`
+
+### Local Search (Brave)
+\`\`\`
+brave_local_search
+- query: string (e.g., "coffee shops near Seattle")
+- count?: number
+- Returns: local business results with:
+  - name, address, phone
+  - rating, reviews
+  - hours, distance
+\`\`\`
+
+## Search Query Optimization
+
+### Effective Queries
+\`\`\`
+# Be specific
+Good: "Next.js 14 app router middleware authentication"
+Bad: "nextjs auth"
+
+# Include version when relevant
+Good: "React 18 useEffect cleanup"
+Bad: "react useEffect"
+
+# Use quotes for exact phrases
+"error boundary" React 18
+
+# Include context
+"TypeScript strict mode" tsconfig.json setup
+\`\`\`
+
+### Query Patterns by Use Case
+\`\`\`
+# Technical documentation
+"[library] [version] official documentation [topic]"
+Example: "Prisma 5 official documentation migrations"
+
+# Error solutions
+"[exact error message]" [framework] [version]
+Example: "Cannot find module 'next/headers'" Next.js 14
+
+# Comparisons
+"[option A] vs [option B]" [year]
+Example: "Zustand vs Jotai 2024"
+
+# Best practices
+"[technology] best practices [year]"
+Example: "TypeScript best practices 2024"
+
+# Breaking changes
+"[library] [version] breaking changes migration"
+Example: "React Router v6 breaking changes migration"
+\`\`\`
+
+## Search Result Handling
+
+### Evaluating Results
+\`\`\`
+1. Check result age - prefer recent for fast-moving tech
+2. Prioritize official docs over third-party
+3. Look for multiple sources confirming same info
+4. Consider Stack Overflow answers with high votes
+5. Check GitHub issues for edge cases
+\`\`\`
+
+### Result Quality Indicators
+\`\`\`
+High quality:
+- Official documentation sites
+- GitHub repositories/issues
+- Stack Overflow (high votes)
+- Recognized tech blogs
+- Recent publication date
+
+Lower quality:
+- Content farms
+- Very old articles for current tech
+- AI-generated content aggregators
+- SEO-optimized tutorials with ads
+\`\`\`
+
+## Common Search Workflows
+
+### Research New Technology
+\`\`\`
+1. Search: "[tech] official documentation"
+2. Search: "[tech] getting started tutorial"
+3. Search: "[tech] vs [alternatives] comparison"
+4. Search: "[tech] production experience"
+5. Synthesize findings
+\`\`\`
+
+### Debug Error
+\`\`\`
+1. Search exact error message in quotes
+2. Add framework/library context
+3. Check GitHub issues if open source
+4. Look for Stack Overflow solutions
+5. Search for related error patterns
+\`\`\`
+
+### Find Best Practices
+\`\`\`
+1. Search: "[topic] best practices 2024"
+2. Search: "[topic] common mistakes avoid"
+3. Search: "[topic] production tips"
+4. Cross-reference multiple sources
+\`\`\`
+
+### API/Library Usage
+\`\`\`
+1. Search: "[library] [method] example"
+2. Search: "[library] [use case] tutorial"
+3. Search: "[library] TypeScript types"
+4. Check official docs for edge cases
+\`\`\`
+
+## Filtering Strategies
+\`\`\`
+# Include specific domains
+include_domains: ["github.com", "stackoverflow.com", "docs.*.com"]
+
+# Exclude low-quality sources
+exclude_domains: ["w3schools.com", "geeksforgeeks.org"]
+
+# Time-sensitive searches
+Add year to query: "React hooks 2024"
+Filter by age in results
+\`\`\`
+
+## Rate Limits & Quotas
+\`\`\`
+Brave Search:
+- Free tier: 2,000 queries/month
+- Check X-RateLimit headers
+
+Tavily:
+- Free tier: 1,000 queries/month
+- Rate limit: 100 requests/minute
+
+Best practices:
+- Cache frequently searched queries
+- Batch related searches
+- Use specific queries to reduce iterations
+\`\`\`
+
+## ❌ DON'T
+- Search for generic terms without context
+- Trust first result blindly
+- Ignore result publication dates
+- Search for sensitive/private information
+- Exceed rate limits with repeated queries
+- Use AI-generated content as primary source
+
+## ✅ DO
+- Use specific, contextual queries
+- Include version numbers for tech
+- Verify information across sources
+- Prefer official documentation
+- Check result freshness for tech topics
+- Use domain filtering for quality
+- Respect rate limits