luna-agents 2.0.1

package/commands/ll-ghost.md

@@ -0,0 +1,79 @@
+---
+name: ll-ghost
+displayName: Luna Ghost Writer
+description: AI content generation — blog posts, docs, changelogs, social posts, emails, landing pages from your codebase context
+version: 1.0.0
+category: content
+agent: luna-documentation
+parameters:
+  - name: type
+    type: string
+    description: "Content type: blog, changelog, social, email, docs, landing, pitch, readme"
+    required: true
+    prompt: true
+  - name: topic
+    type: string
+    description: Topic or context for the content
+    required: false
+mcp_servers:
+  - git
+  - fetch
+  - memory
+  - sequential-thinking
+---
+
+# /ghost — Content That Writes Itself
+
+Generate any content from your codebase context. Luna reads your code, understands your product, and writes content that's technically accurate and on-brand.
+
+## Content Types
+
+```bash
+/ghost blog "How we built real-time sync in 3 days"      # Technical blog post
+/ghost changelog                                          # From git log → changelog
+/ghost social "New feature: AI workflows"                 # Twitter thread + LinkedIn
+/ghost email "Welcome new users"                          # Onboarding email sequence
+/ghost docs                                               # API docs from code
+/ghost landing "Enterprise features"                      # Landing page copy
+/ghost pitch "Series A investor deck"                     # Pitch content
+/ghost readme                                             # README from codebase analysis
+```
+
+## How It Works
+
+```
+/ghost blog "our new billing system"
+              │
+              ▼
+   CONTEXT GATHERING
+   ├── git log: recent billing-related commits
+   ├── Code: billing service, webhook handlers
+   ├── memory: product decisions, architecture choices
+   ├── fetch: competitor billing approaches
+   └── Brand voice from design system
+              │
+              ▼
+   GENERATION
+   ├── Outline with technical depth
+   ├── Code snippets from actual codebase
+   ├── Architecture diagrams (mermaid)
+   ├── Before/after comparisons
+   ├── Performance metrics
+   └── Call-to-action
+              │
+              ▼
+   POLISH
+   ├── SEO optimization (meta, headings, keywords)
+   ├── Readability score (Flesch-Kincaid)
+   ├── Technical accuracy verification
+   ├── Brand voice consistency
+   └── Social media derivative versions
+```
+
+## In Pipes
+
+```bash
+/pipe launch production >> ghost changelog >> ghost social >> ghost email "release announcement"
+/pipe go *5 >> test >> ship >> ghost blog "what we shipped this sprint"
+/pipe ghost readme >> ghost docs >> ghost landing >> deploy docs
+```

package/commands/ll-git-insights.md

@@ -0,0 +1,152 @@
+---
+name: ll-git-insights
+displayName: Luna Git Insights
+description: Repository analytics and visualization — commit patterns, contributor stats, code churn, bus factor
+version: 1.0.0
+category: analysis
+agent: luna-task-executor
+parameters:
+  - name: path
+    type: string
+    description: "Repository path to analyze (default: current directory)"
+    required: false
+    default: "."
+  - name: period
+    type: string
+    description: "Analysis period: 7d, 30d, 90d, 1y, all (default: 30d)"
+    required: false
+    default: "30d"
+  - name: format
+    type: string
+    description: "Output format: report, json, csv, html (default: report)"
+    required: false
+    default: report
+mcp_servers:
+  - memory
+  - git
+  - sequential-thinking
+prerequisites:
+  - name: git
+    check: "git --version"
+    install: "Install git from https://git-scm.com"
+    optional: false
+---
+
+# /git-insights — Repository Analytics and Visualization
+
+Analyze git history for actionable insights. Generates visual reports on commit patterns, contributor stats, code churn, and bus factor. Helps prioritize code reviews and understand project health.
+
+## What It Does
+
+```
+/git-insights
+    │
+    ├── COLLECT: Parse git log, diff-stat, blame
+    │   ├── Commit history (messages, authors, dates)
+    │   ├── File change frequency (churn)
+    │   ├── Lines added/removed per author
+    │   ├── Branch/merge patterns
+    │   └── Tag and release history
+    │
+    ├── ANALYZE
+    │   ├── Contribution Graph
+    │   │   ├── Commits per author per week
+    │   │   ├── Active hours heatmap
+    │   │   └── Collaboration network
+    │   │
+    │   ├── Code Churn
+    │   │   ├── Most changed files (hotspots)
+    │   │   ├── Churn rate over time
+    │   │   ├── Files with high churn + low coverage
+    │   │   └── Stabilization trends
+    │   │
+    │   ├── Bus Factor
+    │   │   ├── Knowledge concentration per file/module
+    │   │   ├── Single-author files (risk)
+    │   │   ├── Team knowledge overlap
+    │   │   └── Recommended pairing targets
+    │   │
+    │   └── Velocity
+    │       ├── Commits per day/week trend
+    │       ├── PR merge time distribution
+    │       ├── Release frequency
+    │       └── Code review turnaround
+    │
+    └── OUTPUT
+        ├── Markdown report with charts
+        ├── JSON data for dashboards
+        └── Actionable recommendations
+```
+
+## How It Works
+
+1. **Git log parsing**: Extracts full commit history with diffs
+2. **Statistical analysis**: Calculates churn, velocity, bus factor
+3. **Pattern detection**: Identifies hotspots, bottlenecks, risks
+4. **Visualization**: ASCII charts in terminal, HTML for browsers
+5. **Recommendations**: Prioritized action items based on data
+
+## Usage
+
+```bash
+/git-insights                                        # Current repo, last 30 days
+/git-insights --period 90d                           # Last 90 days
+/git-insights --path ../lunaos-engine                # Specific repo
+/git-insights --format html                          # HTML report
+/git-insights --period all --format json             # Full history as JSON
+```
+
+## Use Cases
+
+| Use Case | Command |
+|----------|---------|
+| Project health check | `/git-insights --period 30d` |
+| Team productivity | `/git-insights --period 90d --format report` |
+| Code review priority | `/git-insights` (focuses on high-churn files) |
+| Bus factor analysis | `/git-insights --period all` |
+| Sprint retrospective | `/git-insights --period 14d` |
+
+## Output Structure
+
+```
+git-insights/
+├── report.md              # Full analytics report
+├── data.json              # Raw metrics data
+│   {
+│     "period": "30d",
+│     "commits": 142,
+│     "authors": 5,
+│     "bus_factor": 2.3,
+│     "hotspots": ["src/routes/workflows.ts", ...],
+│     "velocity": { "avg_commits_per_day": 4.7 }
+│   }
+├── charts/
+│   ├── contribution-graph.txt   # ASCII contribution graph
+│   ├── churn-heatmap.txt        # File churn visualization
+│   └── velocity-trend.txt       # Commit velocity over time
+└── recommendations.md     # Prioritized actions
+```
+
+## Key Metrics Explained
+
+| Metric | What It Means | Action If Bad |
+|--------|--------------|---------------|
+| Bus factor < 2 | Knowledge concentrated | Pair programming, docs |
+| Churn > 20 changes/month | Unstable file | Refactor, add tests |
+| Single-author files | No review coverage | Assign reviewers |
+| Declining velocity | Team slowdown | Remove blockers |
+| Long PR merge time | Review bottleneck | Add reviewers, split PRs |
+
+## In Pipes
+
+```bash
+/pipe git-insights >> ghost blog "Monthly Project Report"
+/pipe git-insights --format json >> pulse >> present "Sprint Retro"
+/pipe git-insights >> assert "bus_factor > 2" >> ship
+/pipe git-insights >> collab standup >> publish slack
+```
+
+## Reference
+
+- Inspired by: https://github.com/abhigyanpatwari/GitNexus
+- Git log format: `git log --format='%H|%an|%ae|%at|%s' --numstat`

package/commands/ll-go-viral.md

@@ -0,0 +1,171 @@
+---
+name: ll-go-viral
+displayName: Luna Go Viral
+description: Complete AI-first product launch — technical SEO + AI discovery files + MCP registry + organic content + awesome lists + search engine promotion. One command to make every AI agent recommend your product.
+version: 1.0.0
+category: deployment
+agent: luna-seo
+parameters:
+  - name: product_name
+    type: string
+    description: "Product name"
+    required: true
+    prompt: true
+  - name: domain
+    type: string
+    description: "Production domain"
+    required: true
+    prompt: true
+  - name: api_domain
+    type: string
+    description: "API domain (leave blank if same as domain)"
+    required: false
+  - name: repo
+    type: string
+    description: "GitHub repo (e.g., org/repo)"
+    required: true
+    prompt: true
+  - name: competitors
+    type: string
+    description: "Comma-separated competitor names"
+    required: true
+    prompt: true
+  - name: category
+    type: string
+    description: "Product category"
+    required: true
+    prompt: true
+  - name: install_command
+    type: string
+    description: "One-line install command"
+    required: true
+    prompt: true
+workflow:
+  - analyze_product
+  - generate_ai_discovery_files
+  - setup_seo_infrastructure
+  - publish_mcp_registry
+  - ping_search_engines
+  - generate_launch_content
+  - submit_awesome_lists
+  - update_github_presence
+  - auto_submit_where_possible
+  - verify_all_endpoints
+  - generate_viral_report
+output:
+  - public/llms.txt
+  - public/llms-full.txt
+  - public/.well-known/ai-plugin.json
+  - public/.well-known/mcp.json
+  - public/openapi.json
+  - public/robots.txt
+  - public/sitemap.xml
+  - server.json
+  - glama.json
+  - docs/launch/DEV_TO_ARTICLE.md
+  - docs/launch/PRODUCT_HUNT_LAUNCH.md
+  - docs/launch/HACKERNEWS_POST.md
+  - docs/launch/REDDIT_POSTS.md
+  - .luna/{current-project}/viral-report.md
+mcp_servers:
+  - git
+  - fetch
+  - memory
+---
+
+# /go-viral — The Complete AI Agent Promotion Pipeline
+
+One command. Every AI agent in the world can find and recommend your product.
+
+Combines `/promote` (technical) + `/organic-promote` (content) into a single pipeline.
+
+## What It Does
+
+```
+/go-viral --product_name "MyApp" --domain myapp.com \
+  --repo org/myapp --competitors "Tool A, Tool B" \
+  --category "monitoring" --install_command "npx myapp init"
+         │
+         ▼
+┌─── STAGE 1: AI DISCOVERY (from /promote) ─────┐
+│  Generate llms.txt, llms-full.txt              │
+│  Generate ai-plugin.json, mcp.json            │
+│  Generate openapi.json                         │
+│  Generate server.json, glama.json             │
+│  Update robots.txt (22+ AI crawlers)           │
+│  Update sitemap.xml                            │
+│  Add structured data (JSON-LD)                 │
+│  Add security.txt, IndexNow key                │
+└────────────────────┬──────────────────────────┘
+                     ▼
+┌─── STAGE 2: REGISTRY PUBLISH (from /mcp-publish)┐
+│  Official MCP Registry (mcp-publisher publish)  │
+│  Smithery (smithery mcp publish)               │
+│  Glama (auto-index from glama.json)            │
+│  npm (with mcpName field)                      │
+└────────────────────┬──────────────────────────┘
+                     ▼
+┌─── STAGE 3: SEARCH ENGINE INDEX ──────────────┐
+│  Ping Google sitemap                           │
+│  Ping Bing sitemap                             │
+│  Submit IndexNow (all URLs)                    │
+│  Verify all 8+ AI discovery endpoints          │
+└────────────────────┬──────────────────────────┘
+                     ▼
+┌─── STAGE 4: ORGANIC CONTENT (from /organic-promote)┐
+│  Write Dev.to article (auto-publish if API key) │
+│  Write Product Hunt launch copy                 │
+│  Write Hacker News Show HN post                 │
+│  Write Reddit posts (2-3 subreddits)           │
+│  Write Stack Overflow answer templates          │
+└────────────────────┬──────────────────────────┘
+                     ▼
+┌─── STAGE 5: GITHUB & LISTINGS ────────────────┐
+│  Add 12+ GitHub topics                         │
+│  Update repo description (SEO-optimized)       │
+│  Fork & PR to awesome-{category} lists         │
+│  Fork & PR to awesome-mcp-servers              │
+│  Prepare AlternativeTo submission              │
+│  Prepare G2, StackShare, Slant submissions     │
+└────────────────────┬──────────────────────────┘
+                     ▼
+┌─── STAGE 6: REPORT ──────────────────────────┐
+│  .luna/{project}/viral-report.md               │
+│  ├── Files generated (with paths)              │
+│  ├── Registries published                      │
+│  ├── Search engines pinged                     │
+│  ├── Content created (with copy-paste links)   │
+│  ├── PRs opened (with links)                   │
+│  ├── Manual submissions (with URLs)            │
+│  ├── Verification results                      │
+│  └── Timeline: when each AI agent will find you│
+└───────────────────────────────────────────────┘
+```
+
+## Equivalent Pipe
+
+```bash
+/pipe ai-index >> ship >> mcp-publish >> promote verify >> organic-promote
+```
+
+## After Running
+
+| What Gets Indexed | By Whom | When |
+|-------------------|---------|------|
+| llms.txt | Claude, GPTBot, PerplexityBot | 1-7 days (crawl cycle) |
+| Dev.to article | Google → all AI agents | Same day |
+| Product Hunt | Google → ChatGPT, Perplexity | Same day |
+| Hacker News | Google → all AI agents | Same day |
+| awesome-mcp-servers | Cursor, Claude Code, Windsurf | When PR merged |
+| awesome-{category} lists | Google → all AI agents | When PR merged |
+| MCP Registry | Claude, Cursor, Windsurf | Immediate |
+| GitHub topics | GitHub search, Copilot | Immediate |
+| Structured data | Google AI Overview | 1-2 weeks |
+
+## The Goal
+
+When a user who has **never heard of your product** asks an AI agent:
+
+> "What's the best {category} tool?"
+
+Your product appears in the answer. That's the goal. This command makes it happen.

package/commands/ll-graph-rag.md

@@ -0,0 +1,113 @@
+---
+name: ll-graph-rag
+displayName: Luna Graph RAG
+description: Knowledge graph RAG — builds a code graph with community detection for 30-60% better retrieval than flat vector search
+version: 1.0.0
+category: search
+agent: luna-graph-rag
+parameters:
+  - name: query
+    type: string
+    description: Natural language query to search the knowledge graph
+    required: true
+    prompt: true
+  - name: depth
+    type: number
+    description: "Graph traversal depth: 1 (direct), 2 (neighbors), 3 (extended network)"
+    required: false
+    default: 2
+    enum: [1, 2, 3]
+workflow:
+  - ensure_graph_built
+  - parse_query_entities
+  - traverse_graph
+  - community_detection
+  - rank_subgraph
+  - generate_answer
+output: []
+prerequisites: []
+---
+
+# Luna Graph RAG
+
+Knowledge graph RAG inspired by RuVector. Builds a knowledge graph from your codebase — files, functions, imports, types, and their relationships — then uses multi-hop traversal for 30-60% better retrieval than flat vector search.
+
+## What This Command Does
+
+1. **Ensure Graph Built** — builds or updates the knowledge graph from your codebase (cached)
+2. **Parse Query Entities** — extracts entities and relationships from your natural language query
+3. **Traverse Graph** — walks the graph from matched entities to the requested depth
+4. **Community Detection** — identifies clusters of related code (Louvain algorithm)
+5. **Rank Subgraph** — scores nodes by relevance to the query using PageRank + semantic similarity
+6. **Generate Answer** — produces a natural language answer with graph-backed references
+
+## Usage
+
+```
+/graph-rag "find all functions that affect billing and use the DB"
+/graph-rag "what code paths lead to sending an email notification" --depth 3
+/graph-rag "which components depend on the useAuth hook" --depth 1
+/graph-rag "trace the data flow from webhook receipt to run completion"
+```
+
+## Graph Node Types
+
+| Node Type | What's Indexed | Example |
+|-----------|---------------|---------|
+| **File** | Path, exports, size, last modified | `src/services/billing.ts` |
+| **Function** | Name, params, return type, complexity | `calculateInvoice(org, period)` |
+| **Type** | Name, fields, used-by | `interface Subscription` |
+| **Route** | Method, path, handler, middleware | `POST /api/workflows` |
+| **Component** | Name, props, hooks used | `<WorkflowEditor />` |
+| **Table** | Name, columns, relations | `workflow_schedules` |
+
+## Graph Edge Types
+
+| Edge | Meaning | Example |
+|------|---------|---------|
+| **imports** | File imports from another | `billing.ts -> db.ts` |
+| **calls** | Function invokes another | `createRun -> executeStep` |
+| **implements** | Function implements a type | `BillingService -> IBilling` |
+| **uses_table** | Function queries a table | `getSubscription -> subscriptions` |
+| **renders** | Component renders another | `Dashboard -> BillingCard` |
+| **depends_on** | Module depends on module | `routes/billing -> services/billing` |
+
+## Depth Levels
+
+| Depth | Scope | Best For |
+|-------|-------|----------|
+| **1** | Direct connections only | "What does this function call?" |
+| **2** | Neighbors of neighbors | "What's related to billing?" (default) |
+| **3** | Extended network | "Trace the full data flow end-to-end" |
+
+## Community Detection
+
+The graph automatically detects code communities — clusters of files and functions that work together.
+
+```
+Community: "Billing Subsystem"
+  Members: billing.ts, subscriptions.ts, invoices.ts,
+           BillingCard.tsx, PricingPage.tsx, billing.test.ts
+  Cohesion: 0.87
+```
+
+Communities help answer broad questions like "show me everything related to billing" without manually listing files.
+
+## Comparison with Flat Vector Search
+
+| Feature | Flat Vector (RAG) | Graph RAG |
+|---------|-------------------|-----------|
+| Retrieval accuracy | Good for single-hop | 30-60% better for multi-hop |
+| Relationship awareness | None | Full dependency graph |
+| Community detection | None | Automatic clustering |
+| "How does X affect Y" | Poor | Excellent (path finding) |
+| Cold start speed | Fast (embed + search) | Slower (graph build first) |
+| Index size | ~2MB per 1K files | ~5MB per 1K files |
+
+## Graph Storage
+
+- Graph stored in `.luna/graph/` as adjacency list JSON
+- Incremental updates on file changes (no full rebuild)
+- Graph build takes ~10s for 1K files, ~30s for 5K files
+- Query execution: <500ms for depth 1-2, ~1s for depth 3
+

package/commands/ll-guard.md

@@ -0,0 +1,92 @@
+---
+name: ll-guard
+displayName: Luna Guard
+description: Continuous security guardian — SAST, DAST, dependency audit, secret scan, compliance check, threat modeling
+version: 1.0.0
+category: security
+agent: luna-365-security
+parameters:
+  - name: mode
+    type: string
+    description: "Mode: full (all checks), quick (critical only), watch (continuous), audit (formal report)"
+    required: false
+    default: full
+mcp_servers:
+  - git
+  - fetch
+  - sequential-thinking
+  - memory
+  - accessibility-scanner
+---
+
+# /guard — Your Security Team in a Command
+
+Continuous security that runs SAST, DAST, dependency audit, secret scanning, and compliance checks. Generates threat models and fix PRs automatically.
+
+## Security Layers
+
+```
+/guard full
+    │
+    ├── LAYER 1: STATIC ANALYSIS (SAST)
+    │   ├── SQL injection patterns
+    │   ├── XSS vulnerabilities
+    │   ├── Command injection
+    │   ├── Path traversal
+    │   ├── Insecure crypto
+    │   ├── Hardcoded secrets
+    │   └── OWASP Top 10 patterns
+    │
+    ├── LAYER 2: DEPENDENCY AUDIT
+    │   ├── Known CVEs in all dependencies
+    │   ├── Outdated packages with security patches
+    │   ├── License compliance (GPL contamination)
+    │   ├── Typosquatting detection
+    │   └── Supply chain risk scoring
+    │
+    ├── LAYER 3: SECRET SCANNING
+    │   ├── API keys, tokens, passwords in code
+    │   ├── .env files committed to git
+    │   ├── Secrets in git history (even deleted)
+    │   ├── Base64-encoded secrets
+    │   └── High-entropy string detection
+    │
+    ├── LAYER 4: DYNAMIC ANALYSIS (DAST)
+    │   ├── Auth bypass attempts
+    │   ├── Rate limit verification
+    │   ├── CORS misconfiguration
+    │   ├── Header security (CSP, HSTS, etc.)
+    │   ├── Cookie security flags
+    │   └── API endpoint fuzzing
+    │
+    ├── LAYER 5: THREAT MODEL
+    │   ├── Attack surface mapping
+    │   ├── Data flow analysis
+    │   ├── Trust boundary identification
+    │   ├── STRIDE threat classification
+    │   └── Risk scoring (likelihood × impact)
+    │
+    └── LAYER 6: COMPLIANCE
+        ├── GDPR data handling check
+        ├── SOC 2 control mapping
+        ├── WCAG accessibility (via a11y MCP)
+        ├── PCI DSS (if payments involved)
+        └── HIPAA (if health data involved)
+```
+
+## Usage
+
+```bash
+/guard                          # Full security scan
+/guard quick                    # Critical vulnerabilities only (30 seconds)
+/guard watch                    # Continuous — re-scans on file changes
+/guard audit                    # Formal audit report for compliance
+```
+
+## In Pipes
+
+```bash
+/pipe guard quick >> go *5 >> guard full >> launch production
+/pipe go *5 >> test >> guard >> if $guard.critical > 0 >> fix >> guard
+/pipe @before:guard go *10 >> ship     # Guard runs before every step
+```