luna-agents 2.0.1

package/commands/ll-visual-regression.md

@@ -0,0 +1,96 @@
+---
+name: ll-visual-regression
+displayName: Luna Visual Regression
+description: Compare screenshots before/after code changes — detect visual regressions across all pages
+version: 1.0.0
+category: testing
+agent: luna-ui-test
+parameters:
+  - name: url
+    type: string
+    description: App URL to test
+    required: true
+    prompt: true
+  - name: baseline
+    type: string
+    description: Baseline source (previous, branch:main, commit:abc123, url:staging.app.com)
+    required: false
+    prompt: true
+workflow:
+  - load_baseline_screenshots
+  - capture_current_screenshots
+  - pixel_diff_comparison
+  - vision_ai_analysis
+  - generate_regression_report
+output:
+  - .luna/{current-project}/visual-regression/
+  - .luna/{current-project}/visual-regression/report.md
+prerequisites: []
+---
+
+# Luna Visual Regression
+
+Detect visual regressions by comparing screenshots before and after code changes.
+
+## What This Command Does
+
+1. **Baseline** — loads previous screenshots (or captures from another branch/URL)
+2. **Current** — captures fresh screenshots of every page
+3. **Pixel Diff** — compares pixel-by-pixel, highlights differences
+4. **Vision AI** — analyzes diffs to determine if change is intentional or regression
+5. **Report** — visual report with side-by-side comparisons and diff overlays
+
+## Usage
+
+```
+# Compare against last test run
+/visual-regression http://localhost:3000
+
+# Compare against main branch
+/visual-regression http://localhost:3000 branch:main
+
+# Compare against staging
+/visual-regression http://localhost:3000 url:https://staging.myapp.com
+
+# Compare against specific commit
+/visual-regression http://localhost:3000 commit:abc123
+```
+
+## Output
+
+```
+.luna/{project}/visual-regression/
+  report.md                    # Regression report
+  comparisons/
+    landing/
+      baseline.png             # Before
+      current.png              # After
+      diff.png                 # Pixel diff overlay (red = changed)
+      analysis.md              # Vision AI verdict: regression or intentional
+    dashboard/
+      baseline.png
+      current.png
+      diff.png
+      analysis.md
+```
+
+## Diff Thresholds
+
+```yaml
+# In .luna/rules.yaml
+visual_regression:
+  threshold: 0.1               # Max 0.1% pixel difference allowed
+  ignore_antialiasing: true    # Ignore sub-pixel rendering differences
+  ignore_colors: false         # Detect color changes
+  viewports: [375, 768, 1440] # Viewports to compare
+```
+
+## In Pipes
+
+```
+# Visual regression gate before ship
+/pipe refactor >> visual-regression http://localhost:3000 ?>> ship
+
+# Full UI quality pipeline
+/pipe hig >> browser-test >> visual-regression >> a11y >> approve "Ship?" >> ship
+```

package/commands/ll-voice.md

@@ -0,0 +1,138 @@
+---
+name: ll-voice
+displayName: Luna Voice (Voicebox)
+description: Local voice synthesis using Voicebox — open-source ElevenLabs alternative with voice cloning, 23 languages, 5 TTS engines
+version: 2.0.0
+category: creative
+agent: luna-task-executor
+parameters:
+  - name: text
+    type: string
+    description: Text to synthesize into speech
+    required: true
+    prompt: true
+  - name: voice
+    type: string
+    description: "Voice model to use (default: onyx)"
+    required: false
+    default: onyx
+  - name: language
+    type: string
+    description: "Language code — 23 supported (default: en)"
+    required: false
+    default: en
+  - name: output
+    type: string
+    description: "Output format: mp3, wav, ogg, flac (default: mp3)"
+    required: false
+    default: mp3
+mcp_servers:
+  - memory
+  - git
+prerequisites:
+  - name: voicebox
+    check: "curl -s http://localhost:17493/health"
+    install: "See https://github.com/jamiepine/voicebox for setup"
+    optional: true
+---
+
+# /voice — Local Voice Synthesis with Voicebox
+
+Generate speech locally using Voicebox, an open-source ElevenLabs alternative. Clone voices from audio samples, synthesize in 23 languages with 5 TTS engines. Falls back to OpenAI TTS if Voicebox is not running.
+
+## What It Does
+
+```
+/voice "Deploy completed successfully"
+    │
+    ├── CHECK: Voicebox running at localhost:17493?
+    │   ├── YES → Use local Voicebox REST API
+    │   └── NO  → Fall back to OpenAI TTS API
+    │
+    ├── SYNTHESIZE
+    │   ├── Select voice model (onyx, alloy, nova, shimmer, echo, or cloned)
+    │   ├── Apply language settings (23 languages)
+    │   ├── Route through TTS engine (Piper, Coqui, XTTS, VITS, Bark)
+    │   └── Generate audio buffer
+    │
+    └── OUTPUT
+        ├── Save to file (mp3/wav/ogg/flac)
+        ├── Stream to stdout for piping
+        └── Print duration and file path
+```
+
+## How It Works
+
+1. **Local-first**: Voicebox runs as a local REST server at `localhost:17493`
+2. **Voice cloning**: Provide a 10-30 second audio sample to clone any voice
+3. **Multi-engine**: 5 TTS engines — choose quality vs speed tradeoff
+4. **Fallback**: If Voicebox is not running, routes through OpenAI TTS API
+5. **Zero cost**: All local synthesis is free and offline-capable
+
+## Supported Languages
+
+```
+en, es, fr, de, it, pt, nl, pl, ru, zh, ja, ko,
+ar, hi, tr, sv, da, no, fi, cs, el, ro, hu
+```
+
+## Voice Cloning
+
+```bash
+/voice clone --sample ./my-voice-sample.wav    # Create voice clone
+/voice "Hello world" --voice my-clone          # Use cloned voice
+```
+
+Provide a 10-30 second clean audio sample. Voicebox creates a voice profile stored locally. Use it across all synthesis commands.
+
+## Usage
+
+```bash
+/voice "Welcome to LunaOS"                              # Default voice (onyx), English
+/voice "Bonjour le monde" --language fr --voice nova     # French synthesis
+/voice "Deploy complete" --output wav                    # WAV output
+/voice "$(cat README.md)" --voice shimmer                # Narrate your README
+```
+
+## Use Cases
+
+| Use Case | Command |
+|----------|---------|
+| Narrate demos | `/voice "Feature walkthrough" --voice onyx` |
+| Add voice to recordings | `/pipe flow-record >> voice "narration text"` |
+| Podcast intros | `/voice "Welcome to the LunaOS podcast" --voice alloy` |
+| Accessibility audio | `/voice "$(cat docs/guide.md)" --output mp3` |
+| Multi-language demos | `/voice "Hola mundo" --language es` |
+
+## Output Structure
+
+```
+voice-output/
+├── output.mp3          # Generated audio file
+├── metadata.json       # Duration, voice, engine, language
+└── transcript.txt      # Original text input
+```
+
+## In Pipes
+
+```bash
+/pipe ghost blog "topic" >> voice "$content" >> publish            # Blog with audio
+/pipe docs >> voice "$(cat docs/guide.md)" >> record               # Docs narration
+/pipe flow-record "demo" >> voice "walkthrough" >> video            # Demo + narration
+/pipe voice "intro" >> sing "outro" >> publish podcast              # Podcast assembly
+```
+
+## Engine Selection
+
+| Engine | Quality | Speed | VRAM |
+|--------|---------|-------|------|
+| Piper | Good | Fast | None |
+| VITS | Good | Fast | 1 GB |
+| Coqui | Great | Medium | 2 GB |
+| XTTS | Excellent | Slow | 4 GB |
+| Bark | Excellent | Slow | 6 GB |
+
+## Reference
+
+- Voicebox: https://github.com/jamiepine/voicebox
+- Fallback: OpenAI TTS API (requires OPENAI_API_KEY)

package/commands/ll-watch.md

@@ -0,0 +1,65 @@
+---
+name: ll-watch-pipe
+displayName: Luna Watch
+description: Watch file changes and auto-run Luna pipeline on every change
+version: 1.0.0
+category: workflow
+agent: luna-task-executor
+parameters:
+  - name: path
+    type: string
+    description: File path or glob to watch (e.g., src/**/*.ts)
+    required: true
+    prompt: true
+  - name: pipeline
+    type: string
+    description: Luna pipeline to run on change
+    required: true
+    prompt: true
+workflow:
+  - setup_file_watcher
+  - run_pipeline_on_change
+  - report_results
+output: []
+prerequisites: []
+---
+
+# Luna Watch
+
+Auto-run Luna commands when files change.
+
+## Usage
+
+```
+/watch-pipe src/**/*.ts >> test
+/watch-pipe src/components/ >> (test ~~ rev)
+/watch-pipe src/ >> test ?>> ship !>> fix
+/watch-pipe prisma/schema.prisma >> migrate >> test
+```
+
+## How It Works
+
+1. Watches the specified path for file changes
+2. On save/create/delete, runs the Luna pipeline
+3. Debounces rapid changes (waits 500ms after last change)
+4. Shows results inline
+5. Keeps running until you cancel (Ctrl+C)
+
+## Examples
+
+```
+# Auto-test on save
+/watch-pipe src/ >> test
+
+# Auto-review + test on save
+/watch-pipe src/ >> (rev ~~ test)
+
+# Auto-fix accessibility on component change
+/watch-pipe src/components/ >> a11y >> hig
+
+# Auto-migrate on schema change
+/watch-pipe prisma/ >> migrate >> test
+
+# Full quality check on save
+/watch-pipe src/ >> (rev ~~ test ~~ sec) >> log "All passed"
+```

package/commands/ll-workflow.md

@@ -0,0 +1,108 @@
+---
+name: ll-workflow
+displayName: Luna Workflow Manager
+description: Save, load, list, and share named pipeline workflows
+version: 1.0.0
+category: workflow
+agent: luna-task-executor
+parameters:
+  - name: action
+    type: string
+    description: Action (save, load, list, delete, share, templates)
+    required: true
+    prompt: true
+  - name: name
+    type: string
+    description: Workflow name
+    required: false
+    prompt: true
+workflow:
+  - parse_action
+  - execute_workflow_action
+  - update_workflow_registry
+output:
+  - .luna/pipelines/
+prerequisites: []
+---
+
+# Luna Workflow Manager
+
+Save, load, and share reusable pipeline workflows.
+
+## Actions
+
+### Save a workflow
+```
+/workflow save quality-gate "(rev ~~ test ~~ sec ~~ a11y)"
+/workflow save safe-ship "quality-gate ?>> approve 'Ship?' >> ship"
+/workflow save full-dev "req >> des >> plan >> @before:rules @after:test go *10! >> run quality-gate >> ship"
+```
+
+### Load and run a workflow
+```
+/pipe run quality-gate
+/pipe run safe-ship
+/pipe go *5 >> run quality-gate ?>> pr
+```
+
+### List saved workflows
+```
+/workflow list
+```
+
+### Delete a workflow
+```
+/workflow delete old-pipeline
+```
+
+### Templates — pre-built workflows
+```
+/workflow templates
+```
+
+Shows built-in templates you can install:
+
+| Template | Pipeline | Install |
+|----------|----------|---------|
+| `full-dev` | `req >> des >> plan >> go *5 >> rev >> test >> ship >> docs >> watch` | `/workflow save full-dev ...` |
+| `quality-gate` | `(rev ~~ test ~~ sec ~~ a11y)` | `/workflow save quality-gate ...` |
+| `safe-ship` | `quality-gate ?>> approve "Ship?" >> ship !>> fix >> test` | `/workflow save safe-ship ...` |
+| `ai-review` | `search >> nexa review >> nexa bugs >> fix >> test >> pr` | `/workflow save ai-review ...` |
+| `quick-fix` | `(fix >> test) *3? >> pr` | `/workflow save quick-fix ...` |
+| `feature-auto` | `@before:rules @after:test feature >> (rev ~~ sec) ?>> pr` | `/workflow save feature-auto ...` |
+| `brand-launch` | `brand >> auth >> hig >> test >> ship >> docs >> changelog` | `/workflow save brand-launch ...` |
+| `refactor-safe` | `snapshot >> @after:test refactor *3 >> diff >> rev >> pr` | `/workflow save refactor-safe ...` |
+| `multi-deploy` | `map [engine, dashboard, studio] >> in $item (test >> ship)` | `/workflow save multi-deploy ...` |
+| `nightly` | `on schedule:daily >> deps >> sec >> perf >> changelog` | `/workflow save nightly ...` |
+
+## Storage
+
+Workflows are saved to `.luna/pipelines/{name}.pipe`:
+
+```
+.luna/pipelines/
+  quality-gate.pipe
+  safe-ship.pipe
+  full-dev.pipe
+```
+
+Each `.pipe` file contains the pipeline expression and metadata:
+```yaml
+name: quality-gate
+description: Parallel quality checks — review, test, security, accessibility
+created: 2026-03-29
+pipeline: (rev ~~ test ~~ sec ~~ a11y)
+```
+
+## Usage in Pipes
+
+```
+# Reference by name
+/pipe run quality-gate ?>> ship
+
+# Chain multiple workflows
+/pipe run full-dev >> run quality-gate >> run safe-ship
+
+# Use in def
+/pipe def my-flow = go *5 >> run quality-gate >> pr
+```

package/commands/ll-zen.md

@@ -0,0 +1,98 @@
+---
+name: ll-zen
+displayName: Luna Zen
+description: One-command project health restoration — fix ALL issues across code, tests, security, a11y, performance, and deploy
+version: 1.0.0
+category: maintenance
+agent: luna-task-executor
+parameters:
+  - name: scope
+    type: string
+    description: "Scope: all (default), code, tests, security, a11y, perf, deps"
+    required: false
+    default: all
+  - name: aggressive
+    type: boolean
+    description: "Aggressive mode: auto-fix everything without asking (default: false)"
+    required: false
+    default: false
+mcp_servers:
+  - git
+  - accessibility-scanner
+  - zai-mcp-server
+  - playwright
+  - image-compare
+  - fetch
+  - sequential-thinking
+  - memory
+  - ruflo
+---
+
+# /zen — Restore Perfect Project Health
+
+Your project accumulated tech debt, failing tests, security warnings, and accessibility issues? One command to fix everything.
+
+## What /zen Does
+
+```
+/zen
+  │
+  ├── SCAN everything (parallel)
+  │   ├── Linter violations
+  │   ├── TypeScript errors
+  │   ├── Failing tests
+  │   ├── Security vulnerabilities
+  │   ├── Accessibility violations
+  │   ├── Performance regressions
+  │   ├── Outdated dependencies
+  │   ├── Dead code
+  │   ├── Files over 200 lines
+  │   └── Missing documentation
+  │
+  ├── PRIORITIZE by impact
+  │   ├── Critical: security vulns, broken tests
+  │   ├── High: a11y violations, type errors
+  │   ├── Medium: lint warnings, perf issues
+  │   └── Low: style inconsistencies, dead code
+  │
+  ├── FIX (ruflo swarm — parallel agents)
+  │   ├── Agent 1: Fix security issues
+  │   ├── Agent 2: Fix failing tests
+  │   ├── Agent 3: Fix a11y violations
+  │   ├── Agent 4: Update dependencies
+  │   ├── Agent 5: Refactor oversized files
+  │   ├── Agent 6: Remove dead code
+  │   └── Agent 7: Fix TypeScript errors
+  │
+  ├── VERIFY all fixes
+  │   ├── Full test suite passes
+  │   ├── No new TypeScript errors
+  │   ├── Security scan clean
+  │   ├── A11y scan clean
+  │   └── Visual regression check
+  │
+  └── REPORT
+      ├── Issues found: N
+      ├── Issues fixed: M
+      ├── Issues remaining: N-M (with reasons)
+      ├── Before/after health score
+      └── Commit per fix category
+```
+
+## Usage
+
+```bash
+/zen                          # Fix everything
+/zen --scope tests            # Fix all failing tests only
+/zen --scope security         # Fix all security issues only
+/zen --scope deps             # Update all dependencies safely
+/zen --aggressive             # Fix everything without asking
+```
+
+## In Pipes
+
+```bash
+/pipe zen >> test >> browser-test >> launch production
+/pipe zen --aggressive >> pulse >> assert $pulse.score >= 90
+/pipe @before:zen go *10 >> ship    # Zen before every coding session
+```

package/commands/local-llm.md

@@ -0,0 +1,12 @@
+---
+name: local-llm
+displayName: Local LLM (shortcut)
+description: "Shortcut: Run AI locally with llamafile → /ll-local-llm"
+version: 1.0.0
+category: shortcut
+shortcut_for: ll-local-llm
+---
+
+# /local-llm — Shortcut for /ll-local-llm
+
+See /ll-local-llm for full documentation.

package/commands/marketplace.md

@@ -0,0 +1,13 @@
+---
+name: marketplace
+displayName: Marketplace (shortcut)
+description: "Shortcut: Skill marketplace -> /ll-marketplace"
+version: 1.0.0
+category: ecosystem
+shortcut_for: ll-marketplace
+---
+
+# /marketplace — Marketplace
+
+Shortcut for `/ll-marketplace`. Browse, publish, install, and review community skills from the Luna marketplace.
+

package/commands/mcp-publish.md

@@ -0,0 +1,16 @@
+---
+name: mcp-publish
+displayName: MCP Publish (shortcut)
+description: "Shortcut: Publish MCP server to Official Registry, Smithery, Glama -> /ll-mcp-publish"
+version: 1.0.0
+category: deployment
+shortcut_for: ll-mcp-publish
+---
+
+# /mcp-publish — Publish to MCP Registries
+
+Shortcut for `/ll-mcp-publish`. Generate server.json, validate, and publish to Official MCP Registry, Smithery, and Glama.
+
+```bash
+/mcp-publish --name org/project --npm_package my-package --description "My MCP server"
+```

package/commands/migrate.md

@@ -0,0 +1,12 @@
+---
+name: migrate
+displayName: Database Migration (shortcut)
+description: "Shortcut: Generate database migrations -> /ll-migrate"
+version: 1.0.0
+category: database
+shortcut_for: ll-migrate
+---
+
+# /migrate — Database Migration
+
+Shortcut for `/ll-migrate`. Generate safe, reversible database migrations with up/down SQL.

package/commands/money.md

@@ -0,0 +1,12 @@
+---
+name: money
+displayName: Money (shortcut)
+description: "Shortcut: → /ll-money"
+version: 1.0.0
+category: shortcut
+shortcut_for: ll-money
+---
+
+# /money — Shortcut for /ll-money
+
+See /ll-money for full documentation.

package/commands/morph.md

@@ -0,0 +1,12 @@
+---
+name: morph
+displayName: Morph (shortcut)
+description: "Shortcut: → /ll-morph"
+version: 1.0.0
+category: shortcut
+shortcut_for: ll-morph
+---
+
+# /morph — Shortcut for /ll-morph
+
+See /ll-morph for full documentation.

package/commands/multi-agent.md

@@ -0,0 +1,12 @@
+---
+name: multi-agent
+displayName: Multi-Agent (shortcut)
+description: "Shortcut: Run parallel AI agents on isolated branches → /ll-multi-agent"
+version: 1.0.0
+category: shortcut
+shortcut_for: ll-multi-agent
+---
+
+# /multi-agent — Shortcut for /ll-multi-agent
+
+See /ll-multi-agent for full documentation.

package/commands/native.md

@@ -0,0 +1,12 @@
+---
+name: native
+displayName: Native (shortcut)
+description: "Shortcut: → /ll-native"
+version: 1.0.0
+category: shortcut
+shortcut_for: ll-native
+---
+
+# /native — Shortcut for /ll-native
+
+See /ll-native for full documentation.

package/commands/nexa.md

@@ -0,0 +1,12 @@
+---
+name: nexa
+displayName: Nexa AI (shortcut)
+description: "Shortcut: Nexa-powered semantic code analysis -> /ll-nexa"
+version: 1.0.0
+category: ai
+shortcut_for: ll-nexa
+---
+
+# /nexa — Nexa AI
+
+Shortcut for `/ll-nexa`. AI code review, bug detection, architecture recommendations powered by Nexa embeddings.

package/commands/oh.md

@@ -0,0 +1,12 @@
+---
+name: oh
+displayName: OpenHands (shortcut)
+description: "Shortcut: Delegate to OpenHands autonomous agent -> /ll-openhands"
+version: 1.0.0
+category: ai
+shortcut_for: ll-openhands
+---
+
+# /oh — OpenHands
+
+Shortcut for `/ll-openhands`. Delegate complex tasks to the OpenHands autonomous coding agent.

package/commands/organic-promote.md

@@ -0,0 +1,16 @@
+---
+name: organic-promote
+displayName: Organic Promote (shortcut)
+description: "Shortcut: Write articles, submit to awesome lists, Product Hunt, HN, Reddit — make AI agents find you -> /ll-organic-promote"
+version: 1.0.0
+category: deployment
+shortcut_for: ll-organic-promote
+---
+
+# /organic-promote — Get Discovered by AI Agents
+
+Shortcut for `/ll-organic-promote`. Generates Dev.to articles, Product Hunt copy, HN posts, Reddit posts, awesome list PRs, and listing site submissions. Everything AI agents search when users ask questions in your category.
+
+```bash
+/organic-promote --product_name "MyApp" --domain myapp.com --repo org/myapp --competitors "Tool A, Tool B" --category monitoring --install_command "npx myapp init"
+```

package/commands/perf-trace.md

@@ -0,0 +1,12 @@
+---
+name: perf-trace
+displayName: Perf Trace (shortcut)
+description: "Shortcut: Chrome performance tracing with Perfetto → /ll-perf-trace"
+version: 1.0.0
+category: shortcut
+shortcut_for: ll-perf-trace
+---
+
+# /perf-trace — Shortcut for /ll-perf-trace
+
+See /ll-perf-trace for full documentation.

package/commands/perf.md

@@ -0,0 +1,12 @@
+---
+name: perf
+displayName: Performance (shortcut)
+description: "Shortcut: Performance profiling -> /ll-perf"
+version: 1.0.0
+category: quality
+shortcut_for: ll-perf
+---
+
+# /perf — Performance Profiler
+
+Shortcut for `/ll-perf`. Bundle analysis, Core Web Vitals, query optimization, memory leaks.