nemonix-cli 1.0.0

package/src/config/claude-code-config/skills/ralph-tasks/steps/action-add-search.md

@@ -0,0 +1,131 @@
+---
+name: action-add-search
+description: Add a task with codebase context search
+---
+
+# Action: Add Task with Search
+
+Add a task enriched with codebase context for better AI implementation.
+
+## AVAILABLE STATE:
+
+| Variable | Value |
+|----------|-------|
+| `{project_path}` | Current project path |
+| `{tasks_dir}` | Path to .claude/ralph-tasks/ |
+| `{task_description}` | Task description (can be multiple) |
+
+## PRE-CHECK:
+
+Verify {tasks_dir}/tasks.json exists. If not:
+```
+❌ Ralph Tasks not initialized.
+Run: /ralph-tasks setup
+```
+
+## EXECUTION SEQUENCE:
+
+### 1. Detect Multiple Tasks
+
+If input contains multiple tasks (user explicitly lists several):
+- Parse each task separately
+- Process ALL tasks in parallel (see step 2)
+
+### 2. Explore Codebase (PARALLEL)
+
+**For EACH task, launch an Explore agent in parallel:**
+
+```
+Task tool with subagent_type=Explore:
+
+prompt: "Find context for implementing: {task_description}
+
+Search for:
+1. Related files (components, utils, services, routes)
+2. Similar existing implementations to reference
+3. Patterns and conventions used in codebase
+4. Dependencies/imports that might be needed
+5. Test patterns if applicable
+
+Return structured summary:
+- **Relevant files:** paths with line numbers
+- **Patterns:** conventions to follow
+- **Dependencies:** libraries/imports
+- **Reference:** similar code to look at"
+```
+
+**CRITICAL: Launch ALL Explore agents in a SINGLE message for parallel execution.**
+
+### 3. Build Enriched Description
+
+For each task, format:
+```
+{original task description}
+
+## Context
+- **Relevant files:** {discovered files with paths}
+- **Patterns to follow:** {patterns found}
+- **Dependencies:** {packages/imports needed}
+- **Reference:** {similar code paths:line}
+```
+
+### 4. Add Tasks
+
+For each enriched task:
+```bash
+bash ~/.claude/skills/ralph-tasks/scripts/add-task.sh "{project_path}" "{enriched_description}"
+```
+
+### 5. Show Summary
+
+```
+✅ {N} task(s) added with context:
+
+┌─────┬──────────────────────────────────────┐
+│ ID  │ Task                                 │
+├─────┼──────────────────────────────────────┤
+│ #X  │ {task 1 short description}           │
+│ #Y  │ {task 2 short description}           │
+└─────┴──────────────────────────────────────┘
+
+Context discovered:
+- 📁 Task #X: {N} relevant files
+- 📁 Task #Y: {N} relevant files
+
+Run /ralph-tasks list to see all tasks
+Run /ralph-tasks wake-up to start the loop
+```
+
+## MULTI-TASK EXAMPLE:
+
+Input:
+```
+add tasks -s fix the login bug
+add tasks -s add dark mode toggle
+add tasks -s make page load faster
+```
+
+Execution:
+1. Parse 3 tasks
+2. Launch 3 Explore agents IN PARALLEL (single message)
+3. Wait for all to complete
+4. Add 3 enriched tasks
+5. Show summary with all 3
+
+## CONTEXT FORMAT IN TASKS.JSON:
+
+```json
+{
+  "id": 3,
+  "description": "Add dark mode toggle\n\n## Context\n- **Relevant files:** src/components/ThemeProvider.tsx:15-45, src/hooks/useTheme.ts\n- **Patterns to follow:** Uses next-themes, theme in localStorage\n- **Dependencies:** next-themes (installed)\n- **Reference:** ThemeProvider.tsx:15-45 for switching logic",
+  "status": "pending",
+  "createdAt": "2026-01-22T10:00:00Z"
+}
+```
+
+## SUCCESS CRITERIA:
+
+✅ All tasks explored in parallel (single message with multiple Task tools)
+✅ Context enriched for each task
+✅ Tasks added to tasks.json
+✅ Clear summary displayed

package/src/config/claude-code-config/skills/ralph-tasks/steps/action-add.md

@@ -0,0 +1,46 @@
+---
+name: action-add
+description: Add a task to the queue (without search)
+---
+
+# Action: Add Task
+
+Add a single task to the task queue.
+
+## AVAILABLE STATE:
+
+| Variable | Value |
+|----------|-------|
+| `{project_path}` | Current project path |
+| `{tasks_dir}` | Path to .claude/ralph-tasks/ |
+| `{task_description}` | Task description from user |
+
+## PRE-CHECK:
+
+Verify {tasks_dir}/tasks.json exists. If not:
+```
+❌ Ralph Tasks not initialized.
+Run: /ralph-tasks setup
+```
+
+## EXECUTION SEQUENCE:
+
+### 1. Run Add Script
+
+```bash
+bash ~/.claude/skills/ralph-tasks/scripts/add-task.sh "{project_path}" "{task_description}"
+```
+
+### 2. Show Confirmation
+
+```
+✅ Task #{new_id} added: {task_description}
+
+View tasks: /ralph-tasks list
+```
+
+## SUCCESS CRITERIA:
+
+✅ Task added to tasks.json with new ID
+✅ Status set to "pending"
+✅ Confirmation message displayed

package/src/config/claude-code-config/skills/ralph-tasks/steps/action-init.md

@@ -0,0 +1,123 @@
+---
+name: action-init
+description: Parse command and route to appropriate action file
+---
+
+# Action: Initialize & Route
+
+Parse the user's command and route to the appropriate action file.
+
+## EXECUTION SEQUENCE:
+
+### 1. Parse Arguments (Flexible Parsing)
+
+Extract from `$ARGUMENTS`:
+
+**Command detection:**
+- `setup` → setup command
+- `add` or `add tasks` or `add task` → add command
+- `list` or `list tasks` → list command
+- `remove` or `remove task` → remove command
+- `wake-up` or `wakeup` or `wake up` → wake-up command
+
+**Search flag detection:**
+- Look for `--search` or `-s` ANYWHERE in the input
+- Remove the flag from the description
+
+**Task description:**
+- Everything after the command (minus flags) is the description
+- Quotes are OPTIONAL: both `add "task"` and `add task description` work
+- Strip leading/trailing whitespace
+
+**Multi-task detection:**
+- If the input contains multiple separate task descriptions (user explicitly mentions multiple tasks)
+- Process each task separately with parallel Explore agents
+
+### 2. Set State Variables
+
+```
+{project_path} = $(pwd)
+{tasks_dir} = {project_path}/.claude/ralph-tasks
+{command} = parsed command (setup, add, list, remove, wake-up)
+{task_description} = parsed description (for add) - can be unquoted
+{task_id} = parsed ID (for remove)
+{search_mode} = true if --search or -s present
+{multi_task} = true if multiple tasks detected
+```
+
+### 3. Route to Action File
+
+| Command | Condition | Action File |
+|---------|-----------|-------------|
+| `setup` | - | Load `steps/action-setup.md` |
+| `add` | no -s flag | Load `steps/action-add.md` |
+| `add` | with -s/--search | Load `steps/action-add-search.md` |
+| `list` | - | Load `steps/action-list.md` |
+| `remove` | - | Load `steps/action-remove.md` |
+| `wake-up` | - | Load `steps/action-wake-up.md` |
+| (unknown) | - | Show help message |
+
+### 4. Help Message (if no command)
+
+```
+📋 Ralph Tasks - Manage task queue for autonomous AI
+
+Commands:
+  /ralph-tasks setup                    Initialize in current project
+  /ralph-tasks add <description>        Add a task (quotes optional)
+  /ralph-tasks add <description> -s     Add task with codebase context
+  /ralph-tasks list                     List all tasks
+  /ralph-tasks remove <id>              Remove task by ID
+  /ralph-tasks wake-up                  Show loop start command
+
+Examples:
+  /ralph-tasks setup
+  /ralph-tasks add Add user authentication
+  /ralph-tasks add Fix the login bug -s
+  /ralph-tasks add "Add dark mode" --search
+  /ralph-tasks list
+  /ralph-tasks remove 3
+```
+
+## PARSING EXAMPLES:
+
+```
+Input: add Fix the login bug
+→ command=add, task_description="Fix the login bug", search_mode=false
+
+Input: add tasks -s Make the page load faster
+→ command=add, task_description="Make the page load faster", search_mode=true
+
+Input: add "Add authentication" --search
+→ command=add, task_description="Add authentication", search_mode=true
+
+Input: add -s please make this work better
+→ command=add, task_description="please make this work better", search_mode=true
+
+Input: remove 3
+→ command=remove, task_id=3
+```
+
+## MULTI-TASK HANDLING:
+
+If user provides multiple tasks in one message (explicitly numbered or separated):
+1. Parse each task description
+2. For each task with search mode, launch parallel Explore agents
+3. Add all tasks sequentially after exploration completes
+
+Example input:
+```
+add tasks -s fix the login bug
+add tasks -s add dark mode
+add tasks -s improve performance
+```
+
+→ Launch 3 parallel Explore agents, then add 3 tasks.
+
+## SUCCESS CRITERIA:
+
+✅ Command correctly parsed (flexible format accepted)
+✅ Description extracted (with or without quotes)
+✅ Search flag detected anywhere in input
+✅ Multi-task inputs handled
+✅ Correct action file loaded

package/src/config/claude-code-config/skills/ralph-tasks/steps/action-list.md

@@ -0,0 +1,58 @@
+---
+name: action-list
+description: Display all tasks with status
+---
+
+# Action: List Tasks
+
+Display all tasks in the queue with their status.
+
+## AVAILABLE STATE:
+
+| Variable | Value |
+|----------|-------|
+| `{project_path}` | Current project path |
+| `{tasks_dir}` | Path to .claude/ralph-tasks/ |
+
+## PRE-CHECK:
+
+Verify {tasks_dir}/tasks.json exists. If not:
+```
+❌ Ralph Tasks not initialized.
+Run: /ralph-tasks setup
+```
+
+## EXECUTION SEQUENCE:
+
+### 1. Run List Script
+
+```bash
+bash ~/.claude/skills/ralph-tasks/scripts/list-tasks.sh "{project_path}"
+```
+
+### 2. Output Format
+
+```
+╔════════════════════════════════════════════════════════════╗
+║                    📋 RALPH TASKS                          ║
+╚════════════════════════════════════════════════════════════╝
+
+  ✅ #1   [completed]   Add user authentication
+  🔄 #2   [in_progress] Create dashboard page
+  ⏳ #3   [pending]     Add dark mode toggle
+  ⏳ #4   [pending]     Implement notifications
+
+─────────────────────────────────────────────────────────────
+  Summary: 2 pending | 1 in progress | 1 completed
+```
+
+### Status Icons:
+- ⏳ `pending` - Not started
+- 🔄 `in_progress` - Currently being worked on
+- ✅ `completed` - Done
+
+## SUCCESS CRITERIA:
+
+✅ All tasks displayed with ID, status, description
+✅ Status icons correctly shown
+✅ Summary count displayed

package/src/config/claude-code-config/skills/ralph-tasks/steps/action-remove.md

@@ -0,0 +1,48 @@
+---
+name: action-remove
+description: Remove a task by ID
+---
+
+# Action: Remove Task
+
+Remove a task from the queue by its ID.
+
+## AVAILABLE STATE:
+
+| Variable | Value |
+|----------|-------|
+| `{project_path}` | Current project path |
+| `{tasks_dir}` | Path to .claude/ralph-tasks/ |
+| `{task_id}` | ID of task to remove |
+
+## PRE-CHECK:
+
+1. Verify {tasks_dir}/tasks.json exists
+2. Verify {task_id} is a valid number
+3. Verify task with {task_id} exists
+
+Errors:
+```
+❌ Ralph Tasks not initialized. Run: /ralph-tasks setup
+❌ Invalid task ID. Must be a number.
+❌ Task #{task_id} not found.
+```
+
+## EXECUTION SEQUENCE:
+
+### 1. Run Remove Script
+
+```bash
+bash ~/.claude/skills/ralph-tasks/scripts/remove-task.sh "{project_path}" "{task_id}"
+```
+
+### 2. Show Confirmation
+
+```
+✅ Task #{task_id} removed: {task_description}
+```
+
+## SUCCESS CRITERIA:
+
+✅ Task removed from tasks.json
+✅ Confirmation with task description shown

package/src/config/claude-code-config/skills/ralph-tasks/steps/action-setup.md

@@ -0,0 +1,46 @@
+---
+name: action-setup
+description: Initialize ralph-tasks in the current project
+---
+
+# Action: Setup
+
+Initialize the ralph-tasks infrastructure in the current project.
+
+## AVAILABLE STATE:
+
+| Variable | Value |
+|----------|-------|
+| `{project_path}` | Current project path |
+| `{tasks_dir}` | Path to .claude/ralph-tasks/ |
+
+## EXECUTION SEQUENCE:
+
+### 1. Run Setup Script
+
+```bash
+bash ~/.claude/skills/ralph-tasks/scripts/setup.sh "{project_path}"
+```
+
+### 2. Show Success Message
+
+```
+✅ Ralph Tasks initialized!
+
+Created: {tasks_dir}/
+├── tasks.json       # Task queue (empty)
+├── wake-up.sh       # Main loop script
+├── prompt.md        # Agent instructions
+└── progress.txt     # Learnings log
+
+Next steps:
+1. Add tasks: /ralph-tasks add "your task description"
+2. List tasks: /ralph-tasks list
+3. Start loop: sh {tasks_dir}/wake-up.sh
+```
+
+## SUCCESS CRITERIA:
+
+✅ Directory created at {tasks_dir}
+✅ All files created (tasks.json, wake-up.sh, prompt.md, progress.txt)
+✅ Success message displayed

package/src/config/claude-code-config/skills/ralph-tasks/steps/action-wake-up.md

@@ -0,0 +1,62 @@
+---
+name: action-wake-up
+description: Show instructions to start the Ralph loop
+---
+
+# Action: Wake Up
+
+Show the user how to start the Ralph autonomous loop.
+
+## AVAILABLE STATE:
+
+| Variable | Value |
+|----------|-------|
+| `{project_path}` | Current project path |
+| `{tasks_dir}` | Path to .claude/ralph-tasks/ |
+
+## PRE-CHECK:
+
+Verify {tasks_dir}/wake-up.sh exists. If not:
+```
+❌ Ralph Tasks not initialized.
+Run: /ralph-tasks setup
+```
+
+## EXECUTION SEQUENCE:
+
+### 1. Show Instructions
+
+```
+🤖 Start Ralph Loop
+
+Run this command in a terminal:
+
+  sh {tasks_dir}/wake-up.sh
+
+What happens:
+• Processes tasks one by one (lowest ID first)
+• Marks task as in_progress → implements → marks completed
+• Commits changes after each task
+• Waits 60 seconds when no pending tasks
+• Continues indefinitely until Ctrl+C
+
+To stop: Press Ctrl+C
+
+Tips:
+• Run in a separate terminal window
+• Add tasks anytime with: /ralph-tasks add "task"
+• Monitor progress with: /ralph-tasks list
+```
+
+## CRITICAL RULE:
+
+🛑 NEVER run wake-up.sh automatically
+🛑 NEVER execute `sh {tasks_dir}/wake-up.sh` for the user
+
+The user MUST copy and run the command themselves.
+
+## SUCCESS CRITERIA:
+
+✅ Clear instructions displayed
+✅ Command shown for user to copy
+✅ wake-up.sh NOT executed automatically

package/src/config/claude-code-config/skills/utils-fix-errors/SKILL.md

@@ -0,0 +1,61 @@
+---
+name: fix-errors
+description: Fix all ESLint and TypeScript errors with parallel processing using snipper agents
+allowed-tools: Bash(pnpm :*), Bash(tsc :*), Bash(npm :*), Read, Task, Grep
+---
+
+# Fix Errors
+
+Fix all ESLint and TypeScript errors by breaking them into areas and processing in parallel.
+
+## Workflow
+
+1. **DISCOVER COMMANDS**: Check `package.json` for exact script names
+   - Look for: `lint`, `typecheck`, `type-check`, `tsc`, `eslint`, `prettier`, `format`
+
+2. **RUN DIAGNOSTICS**:
+   - Run `pnpm run lint` (or equivalent)
+   - Run `pnpm run typecheck` or `tsc --noEmit`
+   - Capture all error output
+
+3. **ANALYZE ERRORS**:
+   - Extract file paths from error messages
+   - Group errors by file location
+   - Count total errors and affected files
+
+4. **CREATE ERROR AREAS**:
+   - **MAX 5 FILES PER AREA**
+   - Group related files together (same directory/feature)
+   - Example: `Area 1: [file1, file2, file3, file4, file5]`
+
+5. **PARALLEL PROCESSING**: Launch snipper agents for each area
+   - Use Task tool with multiple agents simultaneously
+   - Each agent processes one area (max 5 files)
+   - Provide specific error details for each file
+
+6. **VERIFICATION**: Re-run diagnostics after fixes
+   - Wait for all agents to complete
+   - Re-run lint and typecheck
+   - Report remaining errors
+
+7. **FORMAT CODE**: Apply Prettier (if available)
+   - Run `pnpm run format` or equivalent
+
+## Snipper Agent Instructions
+
+```
+Fix all ESLint and TypeScript errors in these files:
+[list of files with their specific errors]
+
+Focus only on these files. Make minimal changes to fix errors while preserving functionality.
+```
+
+## Rules
+
+- ALWAYS check package.json first for correct commands
+- ONLY fix linting and TypeScript errors
+- NO feature additions - minimal fixes only
+- PARALLEL ONLY - use Task tool for concurrent processing
+- Every error must be assigned to an area
+
+User: $ARGUMENTS

package/src/config/claude-code-config/skills/utils-fix-grammar/SKILL.md

@@ -0,0 +1,59 @@
+---
+name: fix-grammar
+description: Fix grammar and spelling errors in one or multiple files while preserving formatting
+allowed-tools: Read, Edit, Write, MultiEdit, Task
+argument-hint: <file-path> [additional-files...]
+---
+
+# Fix Grammar
+
+Fix grammar and spelling errors in files while preserving formatting and meaning.
+
+## Workflow
+
+1. **PARSE FILES**: Split arguments into individual file paths
+   - **STOP** if no files specified - ask user for file paths
+
+2. **DETERMINE STRATEGY**:
+   - **Single file**: Process directly
+   - **Multiple files**: Launch parallel fix-grammar agents
+
+3. **SINGLE FILE MODE**:
+   - `Read` the file completely
+   - Apply grammar and spelling corrections
+   - `Edit` to update file with corrections
+
+4. **MULTIPLE FILES MODE**:
+   - Use Task tool to launch fix-grammar agent for each file
+   - Process all files simultaneously
+   - Wait for all agents to complete
+
+5. **REPORT**: Show files processed and confirm corrections
+
+## Correction Rules
+
+- Fix ONLY spelling and grammar errors
+- **DO NOT** change meaning or word order
+- **DO NOT** translate anything
+- **DO NOT** modify special tags (MDX, custom syntax, code blocks)
+- **PRESERVE**: All formatting, structure, technical terms
+- Remove any `"""` markers if present
+- Keep the same language used in each sentence
+- Handle multilingual content (keep anglicisms, technical terms)
+
+## Output Format
+
+```
+✓ Fixed grammar in [filename]
+- [number] corrections made
+```
+
+## Rules
+
+- ONLY spelling and grammar corrections
+- PARALLEL processing for multiple files
+- PRESERVE everything: formatting, structure, technical terms
+- MINIMAL changes - corrections only, no improvements
+- Never add explanations or commentary to file content
+
+User: $ARGUMENTS

package/src/config/claude-code-config/skills/utils-oneshot/SKILL.md

@@ -0,0 +1,56 @@
+---
+name: oneshot
+description: Ultra-fast feature implementation using Explore → Code → Test workflow. Use when implementing focused features, single tasks, or when speed over completeness is priority.
+argument-hint: <feature-description>
+---
+
+# OneShot
+
+Implement `$ARGUMENTS` at maximum speed. Ship fast, iterate later.
+
+## Workflow
+
+### 1. EXPLORE (minimal)
+
+Gather minimum viable context:
+- Use `Glob` to find 2-3 key files by pattern
+- Use `Grep` to search for specific patterns
+- Quick `WebSearch` only if library-specific API knowledge needed
+- NO exploration tours - find examples/edit targets and move on
+
+### 2. CODE (main phase)
+
+Execute changes immediately:
+- Follow existing codebase patterns exactly
+- Clear variable/method names over comments
+- Stay STRICTLY in scope - change only what's needed
+- NO comments unless genuinely complex
+- NO refactoring beyond requirements
+- Run formatters if available (`npm run format`)
+
+### 3. TEST (validate)
+
+Check quality:
+- Run: `npm run lint && npm run typecheck` (or equivalent)
+- If fails: fix only what you broke, re-run
+- NO full test suite unless explicitly requested
+
+## Output
+
+When complete, return:
+
+```
+## Done
+
+**Task:** {what was implemented}
+**Files changed:** {list}
+**Validation:** ✓ lint ✓ typecheck
+```
+
+## Constraints
+
+- ONE task only - no tangential improvements
+- NO documentation files unless requested
+- NO refactoring outside immediate scope
+- NO "while I'm here" additions
+- If stuck >2 attempts: report blocker and stop