tech-hub-skills 1.5.1 → 1.5.2

package/.claude/package.json

@@ -1,46 +1,46 @@
-{
-  "name": "tech-hub-skills",
-  "version": "1.5.1",
-  "description": "180+ production-ready AI agent skills for Claude Code and GitHub Copilot",
-  "main": "bin/cli.js",
-  "bin": {
-    "tech-hub-skills": "./bin/cli.js"
-  },
-  "scripts": {
-    "postinstall": "node bin/postinstall.js"
-  },
-  "keywords": [
-    "claude",
-    "claude-code",
-    "github-copilot",
-    "copilot",
-    "ai-agents",
-    "skills",
-    "llm",
-    "ai-engineer",
-    "data-engineer",
-    "security",
-    "enterprise",
-    "vscode"
-  ],
-  "author": "6ogo",
-  "license": "MIT",
-  "repository": {
-    "type": "git",
-    "url": "https://github.com/6ogo/Tech-Skills.git"
-  },
-  "files": [
-    "tech_hub_skills/skills/**/*.md",
-    "tech_hub_skills/roles/**/*.md",
-    "bin/**/*.js",
-    "README.md",
-    "LICENSE"
-  ],
-  "engines": {
-    "node": ">=16.0.0"
-  },
-  "homepage": "https://github.com/6ogo/Tech-Skills",
-  "bugs": {
-    "url": "https://github.com/6ogo/Tech-Skills/issues"
-  }
-}
+{
+  "name": "tech-hub-skills",
+  "version": "1.5.1",
+  "description": "180+ production-ready AI agent skills for Claude Code and GitHub Copilot",
+  "main": "bin/cli.js",
+  "bin": {
+    "tech-hub-skills": "./bin/cli.js"
+  },
+  "scripts": {
+    "postinstall": "node bin/postinstall.js"
+  },
+  "keywords": [
+    "claude",
+    "claude-code",
+    "github-copilot",
+    "copilot",
+    "ai-agents",
+    "skills",
+    "llm",
+    "ai-engineer",
+    "data-engineer",
+    "security",
+    "enterprise",
+    "vscode"
+  ],
+  "author": "6ogo",
+  "license": "MIT",
+  "repository": {
+    "type": "git",
+    "url": "https://github.com/6ogo/Tech-Skills.git"
+  },
+  "files": [
+    "tech_hub_skills/skills/**/*.md",
+    "tech_hub_skills/roles/**/*.md",
+    "bin/**/*.js",
+    "README.md",
+    "LICENSE"
+  ],
+  "engines": {
+    "node": ">=16.0.0"
+  },
+  "homepage": "https://github.com/6ogo/Tech-Skills",
+  "bugs": {
+    "url": "https://github.com/6ogo/Tech-Skills/issues"
+  }
+}

package/.claude/roles/ai-engineer/skills/01-prompt-engineering/README.md

@@ -1,252 +1,252 @@
-# Skill 1: Prompt Engineering & Optimization
-
-## 🎯 Overview
-Master the art and science of crafting, versioning, and optimizing prompts for production LLM applications.
-
-## 🔗 Connections
-- **ML Engineer**: Model evaluation and performance metrics (ml-01)
-- **MLOps**: Prompt versioning and experiment tracking (mo-01, mo-03)
-- **Data Scientist**: A/B testing and statistical analysis of prompt variations (ds-08)
-- **System Design**: Cost optimization and latency management (sd-05)
-- **FinOps**: LLM cost optimization, prompt caching strategies (fo-01, fo-03, fo-07)
-- **DevOps**: CI/CD for prompt templates, version control (do-01, do-05)
-- **Security Architect**: Prompt injection prevention, content safety (sa-08)
-
-## 🛠️ Tools Included
-
-### 1. `prompt_template_manager.py`
-Version-controlled prompt template system with variable injection and inheritance.
-
-### 2. `token_cost_estimator.py`
-Calculate costs across providers (OpenAI, Claude, Gemini) with real-time pricing.
-
-### 3. `prompt_ab_tester.py`
-A/B testing framework for comparing prompt variations with statistical significance.
-
-### 4. `prompt_quality_scorer.py`
-Automated quality scoring for relevance, coherence, and factuality.
-
-## 📊 Key Metrics
-- Token efficiency (output quality per token)
-- Cost per query
-- Response latency
-- Quality scores (0-100)
-
-## 🚀 Quick Start
-
-```python
-from prompt_template_manager import PromptTemplate
-from token_cost_estimator import estimate_cost
-
-# Load a template
-template = PromptTemplate.load("marketing_email_generator")
-
-# Inject variables
-prompt = template.render(product="AI Course", audience="Data Scientists")
-
-# Estimate cost
-cost = estimate_cost(prompt, model="gpt-4", provider="openai")
-print(f"Estimated cost: ${cost:.4f}")
-```
-
-## 📚 Best Practices
-
-### Cost Optimization (FinOps Integration)
-1. **Enable Prompt Caching** - Save up to 90% on costs by caching system prompts and context
-   - Cache static system prompts with `cache_control: ephemeral`
-   - Cache large knowledge bases and conversation history
-   - Monitor cache hit rates and adjust caching strategy
-   - Reference: FinOps fo-07 (AI/ML Cost Optimization)
-
-2. **Track and Optimize Token Usage**
-   - Monitor input/output token ratios
-   - Set token budgets per application/user
-   - Use smaller models (Haiku) for simple tasks, Sonnet/Opus for complex reasoning
-   - Implement token usage alerts and cost dashboards
-   - Reference: FinOps fo-01 (Cost Monitoring), fo-03 (Budget Management)
-
-3. **Optimize Prompt Length**
-   - Remove redundant instructions
-   - Use structured prompts with clear sections
-   - Implement dynamic context pruning for long conversations
-   - Reference: AI Engineer best practices on prompt optimization
-
-### Version Control & Deployment (DevOps Integration)
-4. **Version Prompts with Semantic Versioning**
-   - Store prompts in Git with version tags (v1.0.0, v1.1.0)
-   - Use CI/CD pipelines to deploy prompt changes
-   - Implement blue-green deployments for critical prompts
-   - Reference: DevOps do-01 (CI/CD), do-05 (GitOps)
-
-5. **Automate Prompt Testing**
-   - Run automated tests on prompt changes before deployment
-   - Use golden datasets for regression testing
-   - Implement quality gates in CI/CD pipelines
-   - Reference: DevOps do-02 (Testing Automation)
-
-### Experimentation & Quality (MLOps Integration)
-6. **Use A/B Testing for Production Changes**
-   - Deploy prompt variations to subset of users
-   - Track statistical significance before full rollout
-   - Use experiment tracking (MLflow, Azure ML)
-   - Reference: MLOps mo-01 (Experiment Tracking), Data Scientist ds-08
-
-7. **Monitor Quality Metrics Over Time**
-   - Track quality score degradation (model drift)
-   - Set up alerts for quality drops below thresholds
-   - Implement continuous evaluation pipelines
-   - Reference: MLOps mo-04 (Monitoring), ML Engineer ml-05
-
-### Security & Compliance
-8. **Prevent Prompt Injection Attacks**
-   - Validate and sanitize user inputs
-   - Use structured prompts with clear delimiters
-   - Implement content safety filters
-   - Reference: Security Architect sa-08 (LLM Security)
-
-9. **Audit Prompt Usage**
-   - Log all prompt executions for compliance
-   - Track PII in prompts and responses
-   - Implement GDPR-compliant data retention
-   - Reference: Security Architect sa-01 (PII Detection)
-
-### Azure-Specific Best Practices
-10. **Leverage Azure OpenAI Features**
-    - Use managed identities for authentication
-    - Enable diagnostic logging to Azure Monitor
-    - Implement retry logic with exponential backoff
-    - Use provisioned throughput for high-volume applications
-    - Reference: Azure az-05 (Azure OpenAI Service)
-
-## 💰 Cost Optimization Examples
-
-### Prompt Caching Implementation (90% Cost Savings)
-```python
-from anthropic import Anthropic
-
-client = Anthropic()
-
-# Without caching: $0.015 per request
-# With caching: $0.0015 per request (10x cheaper!)
-
-response = client.messages.create(
-    model="claude-3-5-sonnet-20241022",
-    max_tokens=2048,
-    system=[
-        {
-            "type": "text",
-            "text": "You are a customer support AI assistant with access to our knowledge base...",
-            "cache_control": {"type": "ephemeral"}  # Cache this! Saves 90%
-        },
-        {
-            "type": "text",
-            "text": LARGE_KNOWLEDGE_BASE,  # 50K tokens
-            "cache_control": {"type": "ephemeral"}  # Cache this too!
-        }
-    ],
-    messages=[{"role": "user", "content": user_message}]
-)
-
-# Monitor cache performance
-print(f"Cache hit rate: {response.usage.cache_read_input_tokens / response.usage.input_tokens * 100:.1f}%")
-print(f"Cost savings: ${calculate_savings(response.usage):.4f}")
-```
-
-### Cost Tracking Dashboard
-```python
-from token_cost_estimator import LLMCostTracker
-
-tracker = LLMCostTracker()
-
-# Track all requests
-tracker.log_request(
-    model="claude-3-5-sonnet",
-    input_tokens=1000,
-    output_tokens=500,
-    cached_tokens=800,
-    user_id="team_alpha",
-    project="customer_support"
-)
-
-# Generate cost reports
-monthly_report = tracker.generate_report(period="monthly")
-print(f"Total cost: ${monthly_report.total_cost}")
-print(f"Cost by team: {monthly_report.cost_by_user}")
-print(f"Savings from caching: ${monthly_report.cache_savings}")
-
-# Set budget alerts
-tracker.set_budget_alert(
-    project="customer_support",
-    monthly_budget=1000.00,
-    alert_threshold=0.8  # Alert at 80%
-)
-```
-
-## 🚀 CI/CD for Prompt Templates
-
-### Git-Based Prompt Versioning
-```yaml
-# .github/workflows/prompt-deployment.yml
-name: Deploy Prompts
-
-on:
-  push:
-    paths:
-      - 'prompts/**'
-    branches:
-      - main
-
-jobs:
-  test-and-deploy:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Run prompt tests
-        run: pytest tests/test_prompts.py
-
-      - name: A/B test new prompts
-        run: python scripts/ab_test_prompts.py --canary 10%
-
-      - name: Deploy to production
-        if: success()
-        run: python scripts/deploy_prompts.py --env production
-
-      - name: Monitor quality metrics
-        run: python scripts/monitor_quality.py --duration 1h
-```
-
-## 📊 Enhanced Metrics
-
-| Metric | Description | Target | Monitoring Tool |
-|--------|-------------|--------|-----------------|
-| **Token Efficiency** | Output quality per token | >0.8 | Custom dashboard |
-| **Cost per Query** | Average cost including caching | <$0.01 | Azure Monitor + FinOps dashboard |
-| **Cache Hit Rate** | % of tokens served from cache | >70% | Application Insights |
-| **Response Latency** | P95 latency | <2s | Azure Monitor |
-| **Quality Score** | Automated quality rating | >85/100 | MLOps monitoring |
-| **A/B Test Win Rate** | % of new prompts that beat baseline | >60% | MLflow experiments |
-
-## 🔄 Integration with Other Skills
-
-### End-to-End Workflow
-```
-1. Develop Prompt (ai-01)
-   ↓
-2. Version in Git (do-05)
-   ↓
-3. A/B Test (ds-08, mo-01)
-   ↓
-4. Monitor Costs (fo-01, fo-07)
-   ↓
-5. Deploy via CI/CD (do-01)
-   ↓
-6. Monitor Quality (ml-05, mo-04)
-   ↓
-7. Optimize Caching (fo-07)
-```
-
-## 🎯 Quick Wins
-1. **Enable caching today** - Immediate 70-90% cost reduction for conversational apps
-2. **Set up cost tracking** - Know where your LLM budget is going
-3. **Version prompts in Git** - Enable rollbacks and A/B testing
-4. **Automate testing** - Catch regressions before production
-5. **Monitor quality** - Detect model drift early
+# Skill 1: Prompt Engineering & Optimization
+
+## 🎯 Overview
+Master the art and science of crafting, versioning, and optimizing prompts for production LLM applications.
+
+## 🔗 Connections
+- **ML Engineer**: Model evaluation and performance metrics (ml-01)
+- **MLOps**: Prompt versioning and experiment tracking (mo-01, mo-03)
+- **Data Scientist**: A/B testing and statistical analysis of prompt variations (ds-08)
+- **System Design**: Cost optimization and latency management (sd-05)
+- **FinOps**: LLM cost optimization, prompt caching strategies (fo-01, fo-03, fo-07)
+- **DevOps**: CI/CD for prompt templates, version control (do-01, do-05)
+- **Security Architect**: Prompt injection prevention, content safety (sa-08)
+
+## 🛠️ Tools Included
+
+### 1. `prompt_template_manager.py`
+Version-controlled prompt template system with variable injection and inheritance.
+
+### 2. `token_cost_estimator.py`
+Calculate costs across providers (OpenAI, Claude, Gemini) with real-time pricing.
+
+### 3. `prompt_ab_tester.py`
+A/B testing framework for comparing prompt variations with statistical significance.
+
+### 4. `prompt_quality_scorer.py`
+Automated quality scoring for relevance, coherence, and factuality.
+
+## 📊 Key Metrics
+- Token efficiency (output quality per token)
+- Cost per query
+- Response latency
+- Quality scores (0-100)
+
+## 🚀 Quick Start
+
+```python
+from prompt_template_manager import PromptTemplate
+from token_cost_estimator import estimate_cost
+
+# Load a template
+template = PromptTemplate.load("marketing_email_generator")
+
+# Inject variables
+prompt = template.render(product="AI Course", audience="Data Scientists")
+
+# Estimate cost
+cost = estimate_cost(prompt, model="gpt-4", provider="openai")
+print(f"Estimated cost: ${cost:.4f}")
+```
+
+## 📚 Best Practices
+
+### Cost Optimization (FinOps Integration)
+1. **Enable Prompt Caching** - Save up to 90% on costs by caching system prompts and context
+   - Cache static system prompts with `cache_control: ephemeral`
+   - Cache large knowledge bases and conversation history
+   - Monitor cache hit rates and adjust caching strategy
+   - Reference: FinOps fo-07 (AI/ML Cost Optimization)
+
+2. **Track and Optimize Token Usage**
+   - Monitor input/output token ratios
+   - Set token budgets per application/user
+   - Use smaller models (Haiku) for simple tasks, Sonnet/Opus for complex reasoning
+   - Implement token usage alerts and cost dashboards
+   - Reference: FinOps fo-01 (Cost Monitoring), fo-03 (Budget Management)
+
+3. **Optimize Prompt Length**
+   - Remove redundant instructions
+   - Use structured prompts with clear sections
+   - Implement dynamic context pruning for long conversations
+   - Reference: AI Engineer best practices on prompt optimization
+
+### Version Control & Deployment (DevOps Integration)
+4. **Version Prompts with Semantic Versioning**
+   - Store prompts in Git with version tags (v1.0.0, v1.1.0)
+   - Use CI/CD pipelines to deploy prompt changes
+   - Implement blue-green deployments for critical prompts
+   - Reference: DevOps do-01 (CI/CD), do-05 (GitOps)
+
+5. **Automate Prompt Testing**
+   - Run automated tests on prompt changes before deployment
+   - Use golden datasets for regression testing
+   - Implement quality gates in CI/CD pipelines
+   - Reference: DevOps do-02 (Testing Automation)
+
+### Experimentation & Quality (MLOps Integration)
+6. **Use A/B Testing for Production Changes**
+   - Deploy prompt variations to subset of users
+   - Track statistical significance before full rollout
+   - Use experiment tracking (MLflow, Azure ML)
+   - Reference: MLOps mo-01 (Experiment Tracking), Data Scientist ds-08
+
+7. **Monitor Quality Metrics Over Time**
+   - Track quality score degradation (model drift)
+   - Set up alerts for quality drops below thresholds
+   - Implement continuous evaluation pipelines
+   - Reference: MLOps mo-04 (Monitoring), ML Engineer ml-05
+
+### Security & Compliance
+8. **Prevent Prompt Injection Attacks**
+   - Validate and sanitize user inputs
+   - Use structured prompts with clear delimiters
+   - Implement content safety filters
+   - Reference: Security Architect sa-08 (LLM Security)
+
+9. **Audit Prompt Usage**
+   - Log all prompt executions for compliance
+   - Track PII in prompts and responses
+   - Implement GDPR-compliant data retention
+   - Reference: Security Architect sa-01 (PII Detection)
+
+### Azure-Specific Best Practices
+10. **Leverage Azure OpenAI Features**
+    - Use managed identities for authentication
+    - Enable diagnostic logging to Azure Monitor
+    - Implement retry logic with exponential backoff
+    - Use provisioned throughput for high-volume applications
+    - Reference: Azure az-05 (Azure OpenAI Service)
+
+## 💰 Cost Optimization Examples
+
+### Prompt Caching Implementation (90% Cost Savings)
+```python
+from anthropic import Anthropic
+
+client = Anthropic()
+
+# Without caching: $0.015 per request
+# With caching: $0.0015 per request (10x cheaper!)
+
+response = client.messages.create(
+    model="claude-3-5-sonnet-20241022",
+    max_tokens=2048,
+    system=[
+        {
+            "type": "text",
+            "text": "You are a customer support AI assistant with access to our knowledge base...",
+            "cache_control": {"type": "ephemeral"}  # Cache this! Saves 90%
+        },
+        {
+            "type": "text",
+            "text": LARGE_KNOWLEDGE_BASE,  # 50K tokens
+            "cache_control": {"type": "ephemeral"}  # Cache this too!
+        }
+    ],
+    messages=[{"role": "user", "content": user_message}]
+)
+
+# Monitor cache performance
+print(f"Cache hit rate: {response.usage.cache_read_input_tokens / response.usage.input_tokens * 100:.1f}%")
+print(f"Cost savings: ${calculate_savings(response.usage):.4f}")
+```
+
+### Cost Tracking Dashboard
+```python
+from token_cost_estimator import LLMCostTracker
+
+tracker = LLMCostTracker()
+
+# Track all requests
+tracker.log_request(
+    model="claude-3-5-sonnet",
+    input_tokens=1000,
+    output_tokens=500,
+    cached_tokens=800,
+    user_id="team_alpha",
+    project="customer_support"
+)
+
+# Generate cost reports
+monthly_report = tracker.generate_report(period="monthly")
+print(f"Total cost: ${monthly_report.total_cost}")
+print(f"Cost by team: {monthly_report.cost_by_user}")
+print(f"Savings from caching: ${monthly_report.cache_savings}")
+
+# Set budget alerts
+tracker.set_budget_alert(
+    project="customer_support",
+    monthly_budget=1000.00,
+    alert_threshold=0.8  # Alert at 80%
+)
+```
+
+## 🚀 CI/CD for Prompt Templates
+
+### Git-Based Prompt Versioning
+```yaml
+# .github/workflows/prompt-deployment.yml
+name: Deploy Prompts
+
+on:
+  push:
+    paths:
+      - 'prompts/**'
+    branches:
+      - main
+
+jobs:
+  test-and-deploy:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Run prompt tests
+        run: pytest tests/test_prompts.py
+
+      - name: A/B test new prompts
+        run: python scripts/ab_test_prompts.py --canary 10%
+
+      - name: Deploy to production
+        if: success()
+        run: python scripts/deploy_prompts.py --env production
+
+      - name: Monitor quality metrics
+        run: python scripts/monitor_quality.py --duration 1h
+```
+
+## 📊 Enhanced Metrics
+
+| Metric | Description | Target | Monitoring Tool |
+|--------|-------------|--------|-----------------|
+| **Token Efficiency** | Output quality per token | >0.8 | Custom dashboard |
+| **Cost per Query** | Average cost including caching | <$0.01 | Azure Monitor + FinOps dashboard |
+| **Cache Hit Rate** | % of tokens served from cache | >70% | Application Insights |
+| **Response Latency** | P95 latency | <2s | Azure Monitor |
+| **Quality Score** | Automated quality rating | >85/100 | MLOps monitoring |
+| **A/B Test Win Rate** | % of new prompts that beat baseline | >60% | MLflow experiments |
+
+## 🔄 Integration with Other Skills
+
+### End-to-End Workflow
+```
+1. Develop Prompt (ai-01)
+   ↓
+2. Version in Git (do-05)
+   ↓
+3. A/B Test (ds-08, mo-01)
+   ↓
+4. Monitor Costs (fo-01, fo-07)
+   ↓
+5. Deploy via CI/CD (do-01)
+   ↓
+6. Monitor Quality (ml-05, mo-04)
+   ↓
+7. Optimize Caching (fo-07)
+```
+
+## 🎯 Quick Wins
+1. **Enable caching today** - Immediate 70-90% cost reduction for conversational apps
+2. **Set up cost tracking** - Know where your LLM budget is going
+3. **Version prompts in Git** - Enable rollbacks and A/B testing
+4. **Automate testing** - Catch regressions before production
+5. **Monitor quality** - Detect model drift early