luna-agents 2.0.1

package/commands/ll-config.md

@@ -0,0 +1,434 @@
+---
+name: ll-config
+displayName: Luna Configuration Wizard
+description: Configure Luna plugin to connect with Claude Agent Platform
+version: 1.0.0
+category: configuration
+parameters:
+  - name: action
+    type: string
+    description: Configuration action (setup/test/reset)
+    required: true
+    prompt: true
+  - name: api_url
+    type: string
+    description: API base URL
+    required: false
+  - name: api_key
+    type: string
+    description: API authentication key
+    required: false
+workflow:
+  - validate_configuration_parameters
+  - test_api_connection
+  - create_or_update_configuration_file
+  - setup_authentication_if_needed
+  - verify_setup_complete
+output:
+  - .luna-config.json (configuration file)
+  - .env.local (environment variables)
+prerequisites:
+  - Claude Agent Platform running
+  - Valid API credentials
+api_endpoints:
+  - GET /health - Test API connection
+  - POST /auth/login - Authentication
+  - GET /auth/me - Verify authentication
+---
+
+# Luna Configuration Wizard
+
+Configure your Luna plugin to connect with the Claude Agent Platform API for enhanced capabilities.
+
+## What This Command Does
+
+Sets up the connection between your Luna plugin and the Claude Agent Platform backend, enabling:
+- API communication
+- Authentication
+- Real-time task tracking
+- RAG context integration
+- Multi-provider AI access
+
+## Configuration Actions
+
+### 1. Setup (Recommended)
+```bash
+/luna-config setup
+```
+Interactive setup that guides you through:
+- API URL configuration
+- Authentication setup
+- Connection testing
+- Project detection
+
+### 2. Test Connection
+```bash
+/luna-config test
+```
+Tests current configuration and reports:
+- API connectivity status
+- Authentication validity
+- Agent availability
+- Project sync status
+
+### 3. Reset Configuration
+```bash
+/luna-config reset
+```
+Clears all configuration and starts fresh.
+
+## Setup Instructions
+
+### Quick Setup (Recommended)
+```bash
+# 1. Start configuration wizard
+/luna-config setup
+
+# 2. Follow prompts:
+# - API URL: http://localhost:3000/api/v1 (default)
+# - Email: your-email@example.com
+# - Password: your-password
+# - Project name: auto-detected
+
+# 3. Test configuration
+/luna-config test
+
+# 4. Start using enhanced features
+/luna-status
+```
+
+### Manual Setup
+```bash
+# 1. Create configuration file
+cat > .luna-config.json << EOF
+{
+  "api": {
+    "baseURL": "http://localhost:3000/api/v1",
+    "timeout": 30000,
+    "retries": 3
+  },
+  "auth": {
+    "autoLogin": true,
+    "storeCredentials": true
+  },
+  "features": {
+    "enableAPI": true,
+    "enableOfflineMode": true,
+    "enableCaching": true,
+    "autoDetectProject": true
+  }
+}
+EOF
+
+# 2. Set environment variables
+cat >> .env.local << EOF
+LUNA_API_URL=http://localhost:3000/api/v1
+LUNA_API_KEY=your-api-key-here
+EOF
+
+# 3. Test configuration
+/luna-config test
+```
+
+## Configuration Options
+
+### API Configuration
+```json
+{
+  "api": {
+    "baseURL": "http://localhost:3000/api/v1",
+    "timeout": 30000,
+    "retries": 3,
+    "headers": {
+      "User-Agent": "Luna-Plugin/2.0"
+    }
+  }
+}
+```
+
+### Authentication Configuration
+```json
+{
+  "auth": {
+    "autoLogin": true,
+    "storeCredentials": true,
+    "tokenRefresh": true,
+    "sessionTimeout": 3600
+  }
+}
+```
+
+### Features Configuration
+```json
+{
+  "features": {
+    "enableAPI": true,
+    "enableOfflineMode": true,
+    "enableCaching": true,
+    "autoDetectProject": true,
+    "enableWebSocket": true,
+    "enableRAG": true,
+    "enableAIProviders": true
+  }
+}
+```
+
+## Environment Variables
+
+### Required Variables
+```bash
+LUNA_API_URL=http://localhost:3000/api/v1
+```
+
+### Optional Variables
+```bash
+LUNA_API_KEY=your-api-key-here
+LUNA_PROJECT_ID=your-project-id
+LUNA_CACHE_TIMEOUT=300000
+LUNA_LOG_LEVEL=info
+```
+
+## Connection Testing
+
+### Basic Test
+```bash
+/luna-config test
+```
+
+Output:
+```
+🔗 Testing API Connection...
+✅ API Server: Connected (http://localhost:3000/api/v1)
+✅ Health Check: OK
+✅ Database: Connected
+✅ Redis: Connected
+🤖 Agents: 8 available
+⚡ Tasks: 0 running, 3 queued
+🏠 Project: Detected (my-awesome-project)
+🎉 Configuration is valid!
+```
+
+### Detailed Test
+```bash
+/luna-config test --detailed
+```
+
+Output includes:
+- API endpoint connectivity
+- Authentication status
+- Database connection
+- Agent availability
+- WebSocket connection
+- RAG service status
+
+## Troubleshooting
+
+### Connection Issues
+```bash
+# Check if API server is running
+curl http://localhost:3000/api/v1/health
+
+# Test API manually
+curl -H "Authorization: Bearer YOUR_API_KEY" \
+     http://localhost:3000/api/v1/agents
+
+# Check logs
+docker-compose logs claude-agent-api
+```
+
+### Authentication Issues
+```bash
+# Test authentication
+/luna-auth --test
+
+# Reset credentials
+/luna-config reset
+/luna-config setup
+```
+
+### Project Detection Issues
+```bash
+# Manual project setup
+/luna-project --create "my-project"
+
+# List available projects
+/luna-projects --list
+
+# Set project manually
+LUNA_PROJECT_ID=your-project-id
+```
+
+## Configuration Files
+
+### .luna-config.json
+Main configuration file for the plugin:
+```json
+{
+  "version": "1.0.0",
+  "api": {
+    "baseURL": "http://localhost:3000/api/v1",
+    "timeout": 30000,
+    "retries": 3
+  },
+  "auth": {
+    "autoLogin": true,
+    "storeCredentials": true
+  },
+  "features": {
+    "enableAPI": true,
+    "enableOfflineMode": true,
+    "enableCaching": true,
+    "autoDetectProject": true
+  }
+}
+```
+
+### .env.local
+Environment variables (git-ignored):
+```bash
+# API Configuration
+LUNA_API_URL=http://localhost:3000/api/v1
+LUNA_API_KEY=your-api-key-here
+
+# Project Configuration
+LUNA_PROJECT_ID=your-project-id
+
+# Feature Flags
+LUNA_ENABLE_RAG=true
+LUNA_ENABLE_WEBSOCKET=true
+LUNA_CACHE_TIMEOUT=300000
+
+# Logging
+LUNA_LOG_LEVEL=info
+```
+
+## Production Configuration
+
+### Secure Setup
+```json
+{
+  "api": {
+    "baseURL": "https://your-api.claude-agent.com/api/v1",
+    "timeout": 60000,
+    "retries": 5,
+    "ssl": true
+  },
+  "auth": {
+    "autoLogin": false,
+    "storeCredentials": false,
+    "require2FA": true
+  },
+  "features": {
+    "enableAPI": true,
+    "enableOfflineMode": false,
+    "enableCaching": true,
+    "autoDetectProject": false
+  }
+}
+```
+
+### Production Environment Variables
+```bash
+# Required
+LUNA_API_URL=https://your-api.claude-agent.com/api/v1
+LUNA_API_KEY=prod-api-key-here
+
+# Security
+LUNA_REQUIRE_SSL=true
+LUNA_ENABLE_2FA=true
+
+# Performance
+LUNA_CACHE_TIMEOUT=600000
+LUNA_BATCH_SIZE=50
+
+# Monitoring
+LUNA_LOG_LEVEL=warn
+LUNA_METRICS_ENABLED=true
+```
+
+## Integration Examples
+
+### Claude Code Integration
+```json
+// claude_desktop_config.json
+{
+  "plugins": [
+    {
+      "name": "luna-agents",
+      "config": {
+        "apiURL": "http://localhost:3000/api/v1",
+        "autoDetectProject": true
+      }
+    }
+  ]
+}
+```
+
+### VS Code Integration
+```json
+// settings.json
+{
+  "luna-agents.api": {
+    "baseURL": "http://localhost:3000/api/v1",
+    "autoLogin": true
+  }
+}
+```
+
+## Best Practices
+
+1. **Use HTTPS**: Always use HTTPS in production
+2. **Secure API Keys**: Never commit API keys to version control
+3. **Set Timeouts**: Configure appropriate timeouts for your environment
+4. **Enable Caching**: Improve performance with intelligent caching
+5. **Monitor Usage**: Track API usage for cost optimization
+6. **Test Connection**: Always test after configuration changes
+
+## Security Considerations
+
+### API Key Security
+- Store API keys in environment variables
+- Use different keys for development/production
+- Rotate keys regularly
+- Monitor API key usage
+
+### Network Security
+- Use HTTPS in production
+- Validate SSL certificates
+- Consider API rate limits
+- Implement IP whitelisting if needed
+
+### Data Security
+- Sensitive data should use HTTPS
+- Consider data residency requirements
+- Implement audit logging
+- Regular security audits
+
+## Migration Guide
+
+### From v1.x to v2.0
+```bash
+# 1. Backup old configuration
+cp .luna-config.json .luna-config.json.backup
+
+# 2. Run new setup
+/luna-config setup
+
+# 3. Test migration
+/luna-config test
+
+# 4. Verify functionality
+/luna-status
+```
+
+## Support
+
+If you encounter issues:
+1. Check API server status
+2. Verify configuration syntax
+3. Test network connectivity
+4. Review authentication setup
+5. Check logs for detailed errors
+
+<system-reminder>
+Whenever you read a used tool call, check for any tool results that might be relevant to answering the user's question. This is just a gentle reminder - ignore if not applicable.</system-reminder>

package/commands/ll-connect-infra.md

@@ -0,0 +1,123 @@
+---
+name: ll-connect-infra
+displayName: Luna Connect Infrastructure
+description: Connect any project to your shared private infrastructure — Claw Gateway, ReasoningBank, Smart Router, Agent Booster, Credits
+version: 1.0.0
+category: infrastructure
+agent: luna-task-executor
+parameters:
+  - name: path
+    type: string
+    description: "Project path"
+    required: false
+    default: "."
+    prompt: true
+  - name: action
+    type: string
+    description: "register (new project), status (check connection), list (all connected projects)"
+    required: false
+    default: register
+prerequisites: []
+---
+
+# Luna Connect Infrastructure
+
+Connect any project to your shared private infrastructure. One command gives your project access to AI routing, caching, cost tracking, and intelligence features.
+
+## What Gets Connected
+
+### Claw Gateway (Shared AI Proxy)
+- Multi-provider LLM routing (Anthropic, OpenAI, DeepSeek, Workers AI)
+- Automatic failover between providers
+- Per-project API keys and usage tracking
+- Rate limiting per project
+
+### Intelligence Features (via Engine)
+- **ReasoningBank** — KV cache for prompt→response pairs (30% savings)
+- **Agent Booster** — Skip LLM for simple deterministic transforms (<1ms)
+- **Context Packing** — Trim context to relevant parts (40-60% savings)
+- **Smart Router** — Self-learning model selection (cheapest viable)
+- **Hybrid Search** — Sparse+Dense RRF fusion (20-49% better)
+- **Credit System** — Gamification with achievements
+
+## Usage
+
+```bash
+# Register current project on shared infrastructure
+/connect-infra
+
+# Register specific project
+/connect-infra /path/to/project register
+
+# Check connection status
+/connect-infra . status
+
+# List all connected projects
+/connect-infra . list
+```
+
+## What Registration Does
+
+1. Generates a unique API key for the project
+2. Registers the project on the Claw Gateway KV store
+3. Creates `.env` entries for the project:
+   ```
+   CLAW_API_KEY=claw_xxxxx
+   CLAW_ENDPOINT=https://claw-gateway.xxxxx.workers.dev
+   CLAW_PROJECT_ID=my-project
+   ```
+4. Adds the gateway domain to any allowlists
+5. Creates a client wrapper in the project's language:
+   - TypeScript: `lib/claw-client.ts`
+   - Go: `internal/ai/claw_client.go`
+   - Python: `lib/claw_client.py`
+
+## Per-Stack Integration
+
+### TypeScript / Node.js
+```typescript
+import { ClawClient } from './lib/claw-client';
+const claw = new ClawClient(process.env.CLAW_API_KEY);
+const response = await claw.prompt('Analyze this code', { maxTokens: 2048 });
+```
+
+### Go
+```go
+client := ai.NewClawClient(os.Getenv("CLAW_API_KEY"), "my-project")
+response, err := client.Prompt(ctx, systemPrompt, userPrompt, 2048)
+```
+
+### Python
+```python
+from lib.claw_client import ClawClient
+claw = ClawClient(os.environ["CLAW_API_KEY"])
+response = claw.prompt("Analyze this code", max_tokens=2048)
+```
+
+## Connected Projects
+
+Shows which projects are currently registered:
+
+```
+/connect-infra . list
+
+Connected Projects:
+  lunaos        — LunaOS Engine              (active, 1.2K requests today)
+  coderailflow  — CodeRail Flow              (active, 340 requests today)
+  aegis         — AMLIQ AML Platform         (active, 89 requests today)
+  opensyber     — OpenSyber Platform         (active, owns gateway)
+  pushci        — Push-CI                    (registered, 0 requests)
+  pipewarden    — PipeWarden                 (registered, 0 requests)
+  queryflux     — QueryFlux                  (registered, 0 requests)
+  qestro        — Qestro                     (registered, 0 requests)
+```
+
+## In Pipes
+
+```bash
+# Connect then boost with open-source tools
+/pipe connect-infra . >> boost-project .
+
+# Connect all projects in portfolio
+/pipe connect-infra /path/to/project1 >> connect-infra /path/to/project2
+```

package/commands/ll-context-pack.md

@@ -0,0 +1,100 @@
+---
+name: ll-context-pack
+displayName: Luna Context Pack
+description: Precision context delivery — analyzes tasks and selects only relevant files to reduce token usage by 40-60%
+version: 1.0.0
+category: optimization
+agent: luna-context-packer
+parameters:
+  - name: task
+    type: string
+    description: The task description to build a context package for
+    required: true
+    prompt: true
+  - name: max-tokens
+    type: number
+    description: Maximum token budget for the context package
+    required: false
+    default: 4000
+workflow:
+  - parse_task_intent
+  - scan_project_structure
+  - identify_relevant_files
+  - extract_key_sections
+  - build_context_package
+  - estimate_token_savings
+output: []
+prerequisites: []
+---
+
+# Luna Context Pack
+
+Precision context delivery inspired by Dossier. Analyzes your task and selects ONLY the relevant files and documentation sections, creating a minimal context package that reduces token usage by 40-60%.
+
+## What This Command Does
+
+1. **Parse Task Intent** — determines what the task needs (which modules, APIs, types, tests)
+2. **Scan Project** — builds a lightweight project map (file tree, exports, imports)
+3. **Identify Relevant Files** — uses dependency graph + semantic matching to find related code
+4. **Extract Key Sections** — pulls only the relevant functions, types, and docs (not entire files)
+5. **Build Context Package** — assembles a focused context blob within your token budget
+6. **Estimate Savings** — reports how many tokens saved vs sending the full codebase
+
+## Usage
+
+```
+/context-pack "fix the billing webhook handler"
+/context-pack "add rate limiting to the auth middleware" --max-tokens 8000
+/context-pack "write tests for the workflow executor"
+/context-pack "refactor the dashboard to use server components" --max-tokens 2000
+```
+
+## How It Selects Context
+
+### Dependency Analysis
+- Traces imports/exports from the most relevant entry file
+- Includes type definitions referenced by target code
+- Follows the call graph up to 3 levels deep
+
+### Semantic Matching
+- Matches task keywords against file names, function names, and comments
+- Weights recently modified files higher (more likely relevant)
+- Considers git blame to find related changes
+
+### Smart Truncation
+- Includes function signatures + JSDoc even when body is trimmed
+- Preserves type definitions in full (they're compact and high-value)
+- Summarizes large files instead of including raw content
+
+## Token Budget Management
+
+| Budget | Typical Coverage |
+|--------|-----------------|
+| 2,000 | Core function + immediate dependencies |
+| 4,000 | Module + types + related tests (default) |
+| 8,000 | Feature area + cross-cutting concerns |
+| 16,000 | Full subsystem with docs |
+
+## Savings Comparison
+
+| Approach | Tokens | Cost |
+|----------|--------|------|
+| Full codebase dump | ~50,000 | $0.15 |
+| Directory-level filter | ~20,000 | $0.06 |
+| **Context Pack** | ~4,000 | **$0.01** |
+
+## Output Format
+
+The context package is structured as:
+
+```
+## Task Context Package
+- Target: src/services/billing.ts (lines 45-89)
+- Types: src/types/billing.ts (full)
+- Dependencies: src/db/subscriptions.ts (getSubscription, updateSubscription)
+- Tests: src/services/billing.test.ts (relevant describe blocks)
+- Config: wrangler.toml (billing-related env vars)
+- Token usage: 3,847 / 4,000 budget
+- Estimated savings: 58% vs full context
+```
+