@thanhthi2895/luna-ai 2.16.2

package/agentic/claude/.claude/rules/orchestration-protocol.md

@@ -0,0 +1,43 @@
+# Orchestration Protocol
+
+## Delegation Context (MANDATORY)
+
+When spawning subagents via Task tool, **ALWAYS** include in prompt:
+
+1. **Work Context Path**: The git root of the PRIMARY files being worked on
+2. **Reports Path**: `{work_context}/plans/reports/` for that project
+3. **Plans Path**: `{work_context}/plans/` for that project
+
+**Example:**
+```
+Task prompt: "Fix parser bug.
+Work context: /path/to/project-b
+Reports: /path/to/project-b/plans/reports/
+Plans: /path/to/project-b/plans/"
+```
+
+**Rule:** If CWD differs from work context (editing files in different project), use the **work context paths**, not CWD paths.
+
+---
+
+#### Sequential Chaining
+Chain subagents when tasks have dependencies or require outputs from previous steps:
+- **Planning → Implementation → Simplification → Testing → Review**: Use for feature development (tests verify simplified code)
+- **Research → Design → Code → Documentation**: Use for new system components
+- Each agent completes fully before the next begins
+- Pass context and outputs between agents in the chain
+
+#### Parallel Execution
+Spawn multiple subagents simultaneously for independent tasks:
+- **Code + Tests + Docs**: When implementing separate, non-conflicting components
+- **Multiple Feature Branches**: Different agents working on isolated features
+- **Cross-platform Development**: iOS and Android specific implementations
+- **Careful Coordination**: Ensure no file conflicts or shared resource contention
+- **Merge Strategy**: Plan integration points before parallel execution begins
+
+---
+
+## Agent Teams (Optional)
+
+For multi-session parallel collaboration, activate the `/team` skill.
+Not part of the default orchestration workflow. See `~/.claude/skills/team/SKILL.md` for templates, decision criteria, and spawn instructions.

package/agentic/claude/.claude/rules/primary-workflow.md

@@ -0,0 +1,36 @@
+# Primary Workflow
+
+**IMPORTANT**: Ensure token efficiency while maintaining high quality.
+**IMPORTANT:** Analyze the skills catalog and activate the skills that are needed for the task during the process.
+
+**Routing:**
+- New feature / end-to-end task → `/cook`
+- Bug / error fix → `/fix`
+- Code search / find related implementation paths → `ccc`
+- Broad scouting / quick file discovery → `lu:scout`
+- Visual explanations (diagrams, architecture) → Use Mermaid directly in response
+
+#### 1. Code Implementation
+- Before you start, delegate to `planner` agent to create a implementation plan with TODO tasks in `./plans` directory.
+- When in planning phase, use multiple `researcher` agents in parallel to conduct research on different relevant technical topics and report back to `planner` agent to create implementation plan.
+- For code discovery during planning/implementation, run `ccc search` first; if index is stale or after major refactors/new files, run `ccc index` before search.
+- **[IMPORTANT]** After creating or modifying code file, run compile command/script to check for any compile errors.
+
+#### 2. Testing
+- Delegate to `tester` agent to run tests on the **simplified code**
+- Tests verify the FINAL code that will be reviewed and merged
+- **IMPORTANT:** make sure you don't use fake data, mocks, cheats, tricks, temporary solutions, just to pass the build or github actions.
+- **IMPORTANT:** Always fix failing tests follow the recommendations and delegate to `tester` agent to run tests again, only finish your session when all tests pass.
+
+#### 3. Code Quality
+- After testing passes, delegate to `code-reviewer` agent to review clean, tested code.
+
+#### 4. Integration
+- Always follow the plan given by `planner` agent
+- Document breaking changes
+- Delegate to `docs-manager` agent (Task subagent) to update `./docs` directory if any.
+
+#### 5. Debugging
+- When a user report bugs or issues on the server or a CI/CD pipeline, delegate to `debugger` agent to run tests and analyze the summary report.
+- Read the summary report from `debugger` agent and implement the fix.
+- Delegate to `tester` agent to run tests. If tests fail, fix then repeat from **Step 3**.

package/agentic/claude/.claude/rules/team-coordination-rules.md

@@ -0,0 +1,90 @@
+# Team Coordination Rules
+
+> These rules only apply when operating as a teammate within an Agent Team.
+> They have no effect on standard sessions or subagent workflows.
+
+Rules for agents operating as teammates within an Agent Team.
+
+## File Ownership (CRITICAL)
+
+- Each teammate MUST own distinct files — no overlapping edits
+- Define ownership via glob patterns in task descriptions: `File ownership: src/api/*, src/models/*`
+- Lead resolves ownership conflicts by restructuring tasks or handling shared files directly
+- Tester owns test files only; reads implementation files but never edits them
+- If ownership violation detected: STOP and report to lead immediately
+
+## Git Safety
+
+- Prefer git worktrees for implementation teams — each dev in own worktree eliminates conflicts
+- Never force-push from a teammate session
+- Commit frequently with descriptive messages
+- Pull before push to catch merge conflicts early
+- If working in a git worktree, commit/push to the worktree branch — not main or dev
+
+## Communication Protocol
+
+- Use `SendMessage(type: "message")` for peer DMs — always specify recipient by name
+- Use `SendMessage(type: "broadcast")` ONLY for critical blocking issues affecting entire team
+- Mark tasks completed via `TaskUpdate` BEFORE sending completion message to lead
+- Include actionable findings in messages, not just "I'm done"
+- Never send structured JSON status messages — use plain text
+
+## Luna Stack Conventions
+
+### Report Output
+- Save reports to `{LUNA_REPORTS_PATH}` (injected via hook, fallback: `plans/reports/`)
+- Naming: `{type}-{date}-{slug}.md` where type = your role (researcher, reviewer, debugger)
+- Sacrifice grammar for concision. List unresolved questions at end.
+
+### Commit Messages
+- Use conventional commits: `feat:`, `fix:`, `docs:`, `refactor:`, `test:`, `chore:`
+- No AI references in commit messages
+- Keep commits focused on actual code changes
+
+### Docs Sync (Implementation Teams Only)
+- After completing implementation tasks, lead MUST evaluate docs impact
+- State explicitly: `Docs impact: [none|minor|major]`
+- If impact: update `docs/` directory or note in completion message
+
+## Task Claiming
+
+- Claim lowest-ID unblocked task first (earlier tasks set up context for later ones)
+- Check `TaskList` after completing each task for newly unblocked work
+- Set task to `in_progress` before starting work
+- If all tasks blocked, notify lead and offer to help unblock
+
+## Plan Approval Flow
+
+When `plan_mode_required` is set:
+1. Research and plan your approach (read-only — no file edits)
+2. Send plan via `ExitPlanMode` — this triggers approval request to lead
+3. Wait for lead's `plan_approval_response`
+4. If rejected: revise based on feedback, resubmit
+5. If approved: proceed with implementation
+
+## Conflict Resolution
+
+- If two teammates need the same file: escalate to lead immediately
+- If a teammate's plan is rejected twice: lead takes over that task
+- If findings conflict between reviewers: lead synthesizes and documents disagreement
+- If blocked by another teammate's incomplete work: message them directly first, escalate to lead if unresponsive
+
+## Shutdown Protocol
+
+- Approve shutdown requests unless mid-critical-operation
+- Always mark current task as completed before approving shutdown
+- If rejecting shutdown, explain why concisely
+- Extract `requestId` from shutdown request JSON and pass to `shutdown_response`
+
+## Idle State (Normal Behavior)
+
+- Going idle after sending a message is NORMAL — not an error
+- Idle means waiting for input, not disconnected
+- Sending a message to an idle teammate wakes them up
+- Do not treat idle notifications as completion signals — check task status instead
+
+## Discovery
+
+- Read team config at `~/.claude/teams/{team-name}/config.json` to discover teammates
+- Always refer to teammates by NAME (not agent ID)
+- Names are used for: `recipient` in SendMessage, task `owner` in TaskUpdate

package/agentic/claude/.claude/scripts/set-active-plan.cjs

@@ -0,0 +1,91 @@
+#!/usr/bin/env node
+/**
+ * Update session state with a new active plan.
+ *
+ * Usage: node scripts/set-active-plan.cjs <plan-path>
+ *
+ * This script updates session state when available and always refreshes plans/ACTIVE
+ * so later steps can resolve the latest plan context.
+ */
+
+const fs = require('fs');
+const path = require('path');
+
+// Resolve luna-config-utils path gracefully across different execution contexts
+const possiblePaths = [
+  // 1. If executed globally from ~/.claude/scripts/ (standard case)
+  path.join(__dirname, '../hooks/lib/luna-config-utils.cjs'),
+  
+  // 2. If executed from agentic/shared/scripts/ via symlink pointing to here
+  path.join(__dirname, '../../claude/.claude/hooks/lib/luna-config-utils.cjs'),
+
+  // 3. Fallback for antigravity structure
+  path.join(__dirname, '../../antigravity/.agents/lib/luna-config-utils.cjs')
+];
+
+let utilsPath = null;
+for (const p of possiblePaths) {
+  if (fs.existsSync(p)) {
+    utilsPath = p;
+    break;
+  }
+}
+
+let writeSessionState = () => {};
+let readSessionState = () => null;
+
+if (utilsPath) {
+  try {
+    const utils = require(utilsPath);
+    if (typeof utils.writeSessionState === 'function') writeSessionState = utils.writeSessionState;
+    if (typeof utils.readSessionState === 'function') readSessionState = utils.readSessionState;
+  } catch (err) {
+    console.warn(`Warning: Failed to load luna-config-utils from ${utilsPath}: ${err.message}`);
+  }
+} else {
+  console.warn('Warning: luna-config-utils.cjs not found. Session state sync disabled; only plans/ACTIVE will be updated.');
+}
+
+const sessionId =
+  process.env.CLAUDE_SESSION_ID ||
+  process.env.LUNA_SESSION_ID ||
+  process.env.GEMINI_SESSION_ID ||
+  process.env.OPENCODE_SESSION_ID;
+const newPlan = process.argv[2];
+
+if (!newPlan) {
+  console.error('Error: Plan path required');
+  console.log('Usage: node scripts/set-active-plan.cjs <plan-path>');
+  console.log('Example: node scripts/set-active-plan.cjs plans/260331-1200-feature-name');
+  process.exit(1);
+}
+
+const absolutePlan = path.resolve(newPlan);
+
+if (!sessionId) {
+  console.warn('Warning: SESSION_ID not set - session state will not persist');
+  console.log(`Would set active plan to: ${absolutePlan}`);
+} else {
+  const current = readSessionState(sessionId) || {};
+  writeSessionState(sessionId, {
+    ...current,
+    activePlan: absolutePlan,
+    timestamp: Date.now()
+  });
+}
+
+// Also write to plans/ACTIVE file for backwards compatibility with active-file resolution
+try {
+  const plansDir = path.join(process.cwd(), 'plans');
+  if (!fs.existsSync(plansDir)) {
+    fs.mkdirSync(plansDir, { recursive: true });
+  }
+  const activeFile = path.join(plansDir, 'ACTIVE');
+  fs.writeFileSync(activeFile, absolutePlan, 'utf8');
+  console.log(`Active plan set to: ${absolutePlan}`);
+} catch (err) {
+  console.error(`Failed to write plans/ACTIVE: ${err.message}`);
+  process.exit(1);
+}
+
+process.exit(0);

package/agentic/claude/.claude/settings.json

@@ -0,0 +1,164 @@
+{
+  "env": {
+    "CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS": "1"
+  },
+  "includeCoAuthoredBy": true,
+  "permissions": {},
+  "hooks": {
+    "SessionStart": [
+      {
+        "matcher": "startup|resume|clear|compact",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/session-init.cjs\""
+          }
+        ]
+      }
+    ],
+    "UserPromptSubmit": [
+      {
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/dev-rules-reminder.cjs\""
+          },
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/usage-context-awareness.cjs\"",
+            "timeout": 30
+          }
+        ]
+      }
+    ],
+    "PreToolUse": [
+      {
+        "matcher": "Write",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/descriptive-name.cjs\""
+          }
+        ]
+      },
+      {
+        "matcher": "AskUserQuestion",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/ask-user-notify.cjs\""
+          }
+        ]
+      },
+      {
+        "matcher": "Bash|Glob|Grep|Read|Edit|Write",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/scout-block.cjs\""
+          },
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/privacy-block.cjs\""
+          }
+        ]
+      }
+    ],
+    "PostToolUse": [
+      {
+        "matcher": "Edit|Write|MultiEdit",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/post-edit-reminder.cjs\""
+          }
+        ]
+      },
+      {
+        "matcher": "Bash|Edit|Write|MultiEdit|NotebookEdit",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/usage-context-awareness.cjs\"",
+            "timeout": 10
+          }
+        ]
+      },
+      {
+        "matcher": "Skill",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/skill-loaded-notify.cjs\""
+          }
+        ]
+      }
+    ],
+    "SubagentStart": [
+      {
+        "matcher": ".*",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/subagent-init.cjs\""
+          },
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/team-context-inject.cjs\""
+          }
+        ]
+      }
+    ],
+    "SubagentStop": [
+      {
+        "matcher": "Plan",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/cook-after-plan-reminder.cjs\""
+          }
+        ]
+      }
+    ],
+    "TaskCompleted": [
+      {
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/task-completed-handler.cjs\""
+          }
+        ]
+      }
+    ],
+    "TeammateIdle": [
+      {
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/teammate-idle-handler.cjs\""
+          }
+        ]
+      }
+    ],
+    "Stop": [
+      {
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node \"$HOME/.claude/hooks/completion-notify.cjs\""
+          }
+        ]
+      }
+    ]
+  },
+  "statusLine": {
+    "type": "command",
+    "command": "node \"$HOME/.claude/statusline.cjs\"",
+    "padding": 0
+  },
+  "enabledPlugins": {
+    "rust-analyzer-lsp@claude-plugins-official": true
+  },
+  "skipDangerousModePermissionPrompt": true,
+  "model": "opus[1m]"
+}

package/agentic/claude/.claude/skills/ai-artist/SKILL.md

@@ -0,0 +1,122 @@
+---
+name: lu:ai-artist
+description: "Generate images via Nano Banana with 129 curated prompts. Mandatory validation interview refines style/mood/colors (use --skip to bypass). 3 modes: search, creative, wild. Styles: Ukiyo-e, Bento grid, cyberpunk, cinematic, vintage patent."
+version: 3.1.0
+license: MIT
+argument-hint: "[concept] [--mode search|creative|wild|all] [--skip]"
+---
+
+# AI Artist - Nano Banana Image Generation
+
+Generate images using 129 curated prompts from awesome-nano-banana-pro-prompts collection.
+
+**Validation interview is mandatory** (use `--skip` to bypass).
+
+## Workflow
+
+**IMPORTANT:** Follow `references/validation-workflow.md` when this skill is activated.
+
+## Quick Start
+
+```bash
+python3 scripts/generate.py "<concept>" -o <output.png> [--mode MODE]
+```
+
+### Generation Modes
+
+| Mode | Description |
+|------|-------------|
+| `search` | Find best matching prompt from 129 curated prompts (default) |
+| `creative` | Remix elements from top 3 matching prompts |
+| `wild` | Out-of-the-box creative interpretation (random style transform) |
+| `all` | Generate all 3 variations |
+
+### Examples
+
+```bash
+# Default search mode
+python3 scripts/generate.py "tech conference banner" -o banner.png -ar 16:9
+
+# Creative remix (combines multiple prompts)
+python3 scripts/generate.py "AI workshop" -o workshop.png --mode creative
+
+# Wild/experimental (random artistic transformation)
+python3 scripts/generate.py "product showcase" -o product.png --mode wild
+
+# Generate all 3 variations at once
+python3 scripts/generate.py "futuristic city" -o city.png --mode all -v
+```
+
+### Options
+
+| Flag | Description |
+|------|-------------|
+| `-o, --output` | Output path (required) |
+| `-m, --mode` | search, creative, wild, or all |
+| `-ar, --aspect-ratio` | 1:1, 16:9, 9:16, etc. |
+| `--model` | flash (fast) or pro (quality/4K) |
+| `-v, --verbose` | Show matched prompts and details |
+| `--dry-run` | Show prompt without generating |
+| `--skip` | Bypass validation interview |
+
+---
+
+## Prompt Database
+
+**129 curated prompts** extracted from awesome-nano-banana-pro-prompts:
+
+```bash
+# Search prompts
+python3 scripts/search.py "<query>" --domain awesome
+
+# View all prompts
+cat data/awesome-prompts.csv
+```
+
+### Categories include:
+- **Profile/Avatar**: Thought-leader headshots, mirror selfies
+- **Infographics**: Bento grid, chalkboard, ingredient labels
+- **Social Media**: Quote cards, banners, thumbnails
+- **Product**: Commercial shots, e-commerce, Apple-style
+- **Artistic**: Ukiyo-e, patent documents, vaporwave, cyberpunk
+- **Character**: Anime, chibi, comic storyboards
+
+---
+
+## Wild Mode Transformations
+
+The `wild` mode randomly applies one of these artistic transformations:
+
+- Japanese Ukiyo-e woodblock print
+- Premium liquid glass Bento grid infographic
+- Vintage 1800s patent document
+- Surreal dreamscape with volumetric god rays
+- Cyberpunk neon aesthetic with holograms
+- Hand-drawn chalkboard explanation
+- Isometric 3D diorama
+- Cinematic movie poster
+- Vaporwave aesthetic with glitch effects
+- Apple-style product showcase
+
+---
+
+## References
+
+| Topic | File |
+|-------|------|
+| **Validation Workflow** | `references/validation-workflow.md` |
+| All Prompts | `data/awesome-prompts.csv` |
+| Nano Banana Guide | `references/nano-banana.md` |
+| Image Prompting | `references/image-prompting.md` |
+| Source | `references/awesome-nano-banana-pro-prompts.md` |
+
+---
+
+## Scripts
+
+| Script | Purpose |
+|--------|---------|
+| `generate.py` | Main image generation with 3 modes |
+| `search.py` | Search prompts database |
+| `extract_prompts.py` | Extract prompts from markdown |
+| `core.py` | BM25 search engine |