imhcode 1.0.0

package/.github/ISSUE_TEMPLATE/bug_report.md

@@ -0,0 +1,41 @@
+---
+name: 🐛 Bug Report
+about: Report a bug or unexpected behavior
+title: '[BUG] '
+labels: bug
+assignees: goharabbas321
+---
+
+## Description
+
+A clear and concise description of the bug.
+
+## Steps to Reproduce
+
+1. Run `npx zeoel-framework init`
+2. Open in [AI tool name]
+3. Say "..."
+4. See error
+
+## Expected Behavior
+
+What you expected to happen.
+
+## Actual Behavior
+
+What actually happened. Include error messages or logs if applicable.
+
+## Environment
+
+- **OS**: [e.g., macOS 15, Ubuntu 24.04, Windows 11]
+- **AI Tool**: [e.g., Claude Code, Cursor, Gemini CLI, Copilot, Antigravity]
+- **Zeoel Version**: [e.g., 2.0.1]
+- **Node.js Version**: [e.g., 20.x]
+
+## Screenshots / Logs
+
+If applicable, add screenshots or paste relevant logs.
+
+## Additional Context
+
+Any other context about the problem.

package/.github/ISSUE_TEMPLATE/feature_request.md

@@ -0,0 +1,39 @@
+---
+name: ✨ Feature Request
+about: Suggest a new feature, skill, or agent
+title: '[FEATURE] '
+labels: enhancement
+assignees: ''
+---
+
+## Problem
+
+A clear description of the problem this feature would solve.
+
+> Example: "I'm always frustrated when I have to manually..."
+
+## Proposed Solution
+
+Describe your proposed solution. Be as specific as possible.
+
+## Feature Type
+
+- [ ] New Skill (e.g., `tailwindcss-v4`, `supabase-patterns`)
+- [ ] New Agent (e.g., a specialized DevRel agent)
+- [ ] Pipeline Enhancement (e.g., new phase, better verification)
+- [ ] CLI Improvement (e.g., new `npx zeoel-framework` command)
+- [ ] Documentation Improvement
+- [ ] Other
+
+## Alternatives Considered
+
+Describe any alternative solutions or features you've considered.
+
+## Additional Context
+
+Any other context, mockups, or references that might help.
+
+## Would You Like to Contribute This?
+
+- [ ] Yes, I'd like to submit a PR for this
+- [ ] No, I'm just suggesting it

package/.github/PULL_REQUEST_TEMPLATE.md

@@ -0,0 +1,29 @@
+## Description
+
+Brief description of what this PR does.
+
+## Type of Change
+
+- [ ] 🆕 New skill
+- [ ] 🤖 New agent
+- [ ] 🐛 Bug fix
+- [ ] ✨ Enhancement
+- [ ] 📚 Documentation
+- [ ] 🔧 Maintenance / Refactor
+
+## Changes Made
+
+- 
+- 
+- 
+
+## Checklist
+
+- [ ] My SKILL.md files have proper YAML frontmatter (`name`, `description`)
+- [ ] I tested that the `imhcode` CLI works successfully
+- [ ] I updated `AGENTS.md` if adding a new agent
+- [ ] My changes follow the project conventions
+
+## Related Issues
+
+Fixes #

package/.github/copilot-instructions.md

@@ -0,0 +1,43 @@
+# Copilot Instructions
+
+This project uses the **IMH-Code** framework (Imam Hussain Coding Harness Platform) for multi-agent software development.
+
+## Mandatory Rules (NO EXCEPTIONS)
+
+1. **NEVER** write code without an approved `PROJECT_BRIEF.md`.
+2. **ALWAYS** follow the pipeline: Write (start.md) → Plan (brainstorming.md) → Sprint Plan (plan.md) → Execute → Test.
+3. **ALWAYS** read the relevant agent's manifest and loaded skills before specialized work.
+4. Code **MUST** live in `frontend/` or `backend/`. No other code directories.
+5. When the user says "build X" → Start with Phase 1 (Write & Plan). Do NOT jump to code.
+6. Respect the configured `testing_mode` inside `imhcode.config.json` (fast, balanced, strict).
+
+## Framework Entry Point
+
+Read `skills/imhcode/SKILL.md` at the start of every task.
+
+## Generic Role-Based Agents (19)
+
+| ID | Role | Category |
+|---|---|---|
+| `planner` | Project Planner & Coordinator | planning |
+| `designer` | UX/UI Designer | frontend |
+| `nextjs-executor` | Next.js Full-Stack Developer | frontend |
+| `react-executor` | React/Vite Developer | frontend |
+| `vue-executor` | Vue 3 / Nuxt 4 Developer | frontend |
+| `laravel-executor` | Laravel Backend Developer | backend |
+| `python-executor` | Python Backend (FastAPI/Django) | backend |
+| `java-executor` | Java Backend (Spring Boot) | backend |
+| `flutter-executor` | Flutter Mobile Developer | backend |
+| `react-native-executor` | React Native Mobile Developer | backend |
+| `ios-executor` | iOS Native Developer | backend |
+| `android-executor` | Android Native Developer | backend |
+| `systems-executor` | Systems Developer (Go/Rust/C++) | backend |
+| `web3-executor` | Web3 & Smart Contracts Developer | backend |
+| `tester` | QA & Testing Specialist | testing |
+| `security-auditor` | Cybersecurity Audit Specialist | testing |
+| `seo-optimizer` | SEO Technical Specialist | review |
+| `devops-executor` | DevOps & Cloud Engineer | backend |
+| `debugger` | Systematic Debugging Specialist | review |
+
+All agent files: `agents/[id]/agent.yml` and `SYSTEM.md`.
+All skill files: `skills/[id]/SKILL.md`.

package/.github/workflows/framework-validator.yml

@@ -0,0 +1,53 @@
+name: Zeoel Framework Validator
+
+on:
+  push:
+    branches: [ main ]
+  pull_request:
+    branches: [ main ]
+
+jobs:
+  validate:
+    runs-on: ubuntu-latest
+
+    steps:
+    - name: Checkout repository
+      uses: actions/checkout@v4
+
+    - name: Setup Node.js
+      uses: actions/setup-node@v4
+      with:
+        node-size: 20
+
+    - name: Verify package structure
+      run: |
+        echo "Verifying core files exist..."
+        test -f CLAUDE.md
+        test -f AGENTS.md
+        test -f SETUP.md
+        test -f README.md
+        test -f bin/zeoel.js
+
+    - name: Run Smoke Test (Initialization)
+      run: |
+        echo "Running framework initialization smoke test..."
+        mkdir test-sandbox
+        cd test-sandbox
+        node ../bin/zeoel.js
+        echo "Verifying copied files inside sandbox..."
+        test -d .agents
+        test -d .github
+        test -f CLAUDE.md
+        test -f AGENTS.md
+        test -f SETUP.md
+        test -d frontend
+        test -d backend
+        test -d .worktrees
+        echo "Smoke test passed successfully!"
+
+    - name: Lint Agent and Skill markdown syntax
+      run: |
+        echo "Verifying markdown links and frontmatter formatting..."
+        # Basic check to ensure markdown files aren't completely empty and start cleanly
+        find .agents/skills/zeoel/agents/ -name "*.md" -exec head -n 1 {} \;
+        find .agents/skills/zeoel/skills/ -name "SKILL.md" -exec head -n 1 {} \;

package/.gitignore.template

@@ -0,0 +1,15 @@
+# IMH-Code Framework Ignore Rules
+.agents/
+docs/sprint-*/
+docs/qa/
+docs/security/
+docs/seo/
+docs/tests/
+.worktrees/
+dist/
+node_modules/
+.imhcode/
+imhcode.config.json
+PROJECT_BRIEF.md
+docs/brainstorming.md
+docs/start.md

package/AGENTS.md

@@ -0,0 +1,69 @@
+# Agents
+
+This project uses the **IMH-Code** (Imam Hussain Coding Harness Platform) — a multi-agent orchestration framework for SaaS development.
+
+<HARD-GATE>
+MANDATORY WORKFLOW — NO EXCEPTIONS:
+1. When the user says "build X", "create X", or describes ANY project → Start with Phase 1 (Write) and Phase 2 (Plan). Do NOT write code.
+2. You MUST follow the 6-phase pipeline: Write → Plan → Brainstorm → Sprint Plan → Execute → Test.
+3. Code MUST live in `frontend/` or `backend/`. No other code directories.
+4. You MUST dispatch generic executor agents with their core skill bindings loaded.
+5. You MUST update context.md and PROJECT_BRIEF.md after completing each task.
+</HARD-GATE>
+
+## The Pipeline
+
+| Phase | Description | Deliverables | CLI |
+|-------|-------------|--------------|-----|
+| 1. Write | User inputs requirements | `docs/start.md` | — |
+| 2. Plan | System generates brainstorming questions | `docs/brainstorming.md` | `imhcode plan` |
+| 3. Brainstorm | User answers/reviews recommended answers | `docs/brainstorming.md` (edited) | — |
+| 4. Sprint Plan | System plans roadmaps and task scripts | `PROJECT_BRIEF.md`, `docs/sprint-*/plan.md` | `imhcode plan` |
+| 5. Execute | Sprints executed with intelligent model routing | Working application code | `imhcode execute N` |
+| 6. Test | System audits code, security, SEO, and browser tests | Test reports, security audit, SEO checks | `imhcode test` or final sprint |
+
+## Generic Role-Based Agents (19)
+
+Instead of individual persona names, IMH-Code runs generic agents configured with intelligent model routing:
+
+| Agent ID | Role | Task Category | Primary Skills |
+|---|---|---|---|
+| `planner` | Project Planner & Coordinator | planning | imhcode-core, imhcode-brainstorm, imhcode-sprint-planner |
+| `designer` | UX/UI Designer | frontend | frontend-design, ui-ux-pro-max, modern-web-design |
+| `nextjs-executor` | Next.js Full-Stack Developer | frontend | nextjs-turbopack, shadcn-ui, tailwindcss-v4 |
+| `react-executor` | React/Vite SPA Developer | frontend | vite-patterns, frontend-design |
+| `vue-executor` | Vue 3 / Nuxt 4 Developer | frontend | vue3-composition, nuxt4-patterns |
+| `laravel-executor` | Laravel Backend Developer | backend | laravel-patterns, backend-patterns |
+| `python-executor` | Python Backend (FastAPI/Django) | backend | python-patterns, fastapi-patterns |
+| `java-executor` | Java Backend (Spring Boot) | backend | java-patterns, springboot-patterns |
+| `flutter-executor` | Flutter Mobile Developer | backend | dart-flutter-patterns |
+| `react-native-executor` | React Native Mobile Developer | backend | react-native-best-practices |
+| `ios-executor` | iOS Native Developer | backend | swiftui-patterns |
+| `android-executor` | Android Native Developer | backend | kotlin-patterns |
+| `systems-executor` | Systems Developer (Go/Rust/C++) | backend | golang-patterns, rust-patterns |
+| `web3-executor` | Web3 & Smart Contracts Developer | backend | solidity-patterns |
+| `tester` | QA & Testing Specialist | testing | e2e-testing, webapp-testing |
+| `security-auditor` | Cybersecurity Audit Specialist | testing | imhcode-security, security-scan |
+| `seo-optimizer` | SEO Technical Specialist | review | seo, seo-growth |
+| `devops-executor` | DevOps & Cloud Engineer | backend | deployment-patterns, docker-patterns |
+| `debugger` | Systematic Debugging Specialist | review | systematic-debugging, error-handling |
+
+## Model Routing Matrix
+
+Model routing maps agent categories to primary/fallback models in `imhcode.config.json`:
+
+- **frontend**: Mimo VL v2.5 Pro (Primary) / GPT-5.5 / Claude Sonnet
+- **backend**: DeepSeek V4 Pro (Primary) / GPT-5.5 / Qwen Coder
+- **planning**: Claude Opus 4.6 (Primary) / GPT-5.5
+- **testing**: GPT-5.5 (Primary) / Claude Opus
+- **review**: GPT-5.5 (Primary) / Claude Sonnet
+- **fast**: DeepSeek V4 Flash (Primary) / Gemini Flash
+
+## Sub-Agent Dispatch Protocol
+
+When executing a task:
+1. **Announce** dispatch: "I am now acting as [Agent ID] ([Role])."
+2. **Execute** using bound skills. No TDD by default unless testing_mode is "strict".
+3. **Log** results to `docs/sprint-N/progress.md`.
+4. **Update** `.imhcode/context.md` and `PROJECT_BRIEF.md`.
+5. **Drop** persona: "Returning to Planner."

package/CLAUDE.md

@@ -0,0 +1,27 @@
+# IMH-Code Project Guidelines
+
+## Build Commands
+- Initialize project: `imhcode`
+- Smart planning: `imhcode plan`
+- Run Sprint tasks: `imhcode execute [sprint-number]`
+- Run testing sprint: `imhcode test`
+- Compile typescript orchestrator: `npm run build`
+
+## Codebase Architecture
+- All frontend code goes into `frontend/` directory (Next.js, React, etc.)
+- All backend code goes into `backend/` directory (Laravel, Python, etc.)
+- All plan documents go into `docs/` directory
+- Do not create application code files at the project root
+
+## CLI Model Routing Config
+- Stored in `imhcode.config.json`
+- Maps tasks (frontend, backend, testing, etc.) to optimal LLMs based on available engines
+- Override routing with: `--engine <engine>` and `-m/--model <model>` during manual runs
+
+## Pipeline Workflow (6 Phases)
+1. **Write**: User describes problem/feature requirements in `docs/start.md`.
+2. **Plan**: Run `imhcode plan` to generate `docs/brainstorming.md`.
+3. **Brainstorm**: Review recommended answers in `docs/brainstorming.md`, customize, and run `imhcode plan` again.
+4. **Sprint Plan**: Sprints, tasks, and task script commands are auto-generated.
+5. **Execute**: Run `imhcode execute N`. Respects configured `testing_mode` (fast/balanced/strict).
+6. **Test**: Run `imhcode test` for final testing, security audits, SEO reports, and browser checks.

package/README.md

@@ -0,0 +1,63 @@
+# 🕌 IMH-Code — Imam Hussain Coding Harness Platform
+
+### A fast-first multi-agent AI coding harness with intelligent model routing, configurable testing, and token-saving optimizations for rapid SaaS MVP development.
+
+<p align="center">
+  <a href="https://www.npmjs.com/package/imhcode"><img src="https://img.shields.io/npm/v/imhcode?label=imhcode&style=for-the-badge&logo=npm&color=cb3837" alt="npm version"></a>
+  <a href="https://opensource.org/licenses/MIT"><img src="https://img.shields.io/badge/License-MIT-22c55e.svg?style=for-the-badge" alt="License: MIT"></a>
+  <a href="AGENTS.md"><img src="https://img.shields.io/badge/Agents-19%20Role--Based-a855f7?style=for-the-badge" alt="Agents"></a>
+  <a href="skills/"><img src="https://img.shields.io/badge/Skills-450+%20Bindings-f97316?style=for-the-badge" alt="Skills"></a>
+</p>
+
+---
+
+## ⚡ Quick Start
+
+### 1. Install Globally
+```bash
+npm install -g imhcode
+```
+
+### 2. Initialize a Workspace
+Navigate to your project directory and run the initialization wizard:
+```bash
+cd your-project
+imhcode
+```
+This scans your system for local engines (Claude, OpenCode, Codex, Antigravity, QwenCode, MimoCode) and sets up the **Intelligent Model Routing** configurations inside `imhcode.config.json`.
+
+---
+
+## 🔄 The 6-Phase Pipeline
+
+IMH-Code enforces a fast-first software delivery pipeline:
+
+1. **Write**: Describe your product requirements in `docs/start.md`.
+2. **Plan**: Run `imhcode plan` to generate `docs/brainstorming.md`.
+3. **Brainstorm**: Review recommended answers in `docs/brainstorming.md` and run `imhcode plan` again.
+4. **Sprint Plan**: Sprints, tasks, and shell execution scripts are generated.
+5. **Execute**: Run `imhcode execute 1` (or specify sprint). Respects your selected testing mode.
+6. **Test**: Run `imhcode test` to execute E2E browser tests, security audit, and SEO scans.
+
+---
+
+## 🔒 Configurable Testing Strategy
+
+Unlike other frameworks that mandate slow TDD at every step, IMH-Code lets you choose:
+- **Fast Mode (Default)**: No tests written during development sprints. A dedicated testing sprint is appended at the end to cover unit, integration, security, and browser tests. 3-5x faster development.
+- **Balanced Mode**: Basic smoke tests per task, full test suites at the end.
+- **Strict TDD**: Write failing tests first before coding on every task.
+
+---
+
+## 🤖 19 Generic Role-Based Agents
+
+IMH-Code replaces individual persona names with generic agents routed to the best models:
+- **planner**: Sprint coordination and architecture.
+- **designer**: UI/UX design (prefer Mimo v2.5 Pro).
+- **nextjs-executor**: Next.js full-stack (prefer Mimo v2.5 Pro).
+- **laravel-executor**: Laravel backend (prefer DeepSeek V4 Pro).
+- **tester**: Unit/Integration/E2E tests (prefer GPT-5.5).
+- **security-auditor**: OWASP security scans (prefer GPT-5.5).
+- **seo-optimizer**: SEO & sitemaps (prefer GPT-5.5).
+- *(Run `imhcode agent list` to see all 19 agents)*

package/SETUP.md

@@ -0,0 +1,81 @@
+# IMH-Code Setup Guide — Step by Step
+
+> **Imam Hussain Coding Harness Platform (IMH-Code)**
+> A fast-first multi-agent AI coding harness with intelligent model routing.
+
+---
+
+## Installation & Initialization
+
+1. Install the npm package globally:
+   ```bash
+   npm install -g imhcode
+   ```
+2. Initialize IMH-Code in your project workspace directory:
+   ```bash
+   imhcode
+   ```
+   During initialization, the CLI will:
+   - Scan for local coding assistant CLI engines (Claude, OpenCode, Codex, Antigravity, QwenCode, MimoCode).
+   - Prompt you to configure **intelligent model routing** per task category (frontend, backend, planning, testing, review, fast).
+   - Save the configuration under `imhcode.config.json`.
+   - Setup global configurations under `~/.imhcode/` and rule mappings in assistant directories.
+   - Create local templates (`docs/start.md`).
+
+---
+
+## Project Structure
+
+IMH-Code organizes your codebase into separated root folders:
+
+```
+your-project/
+├── frontend/                  # Next.js / React / Vue web app code
+├── backend/                   # Laravel / Python / Spring Boot API code
+├── docs/                      # Sprints, plans, and testing audits
+├── agents/                    # Manifests & SYSTEM prompts for 19 generic agents
+├── skills/                    # Reusable skill files (SKILL.md)
+├── PROJECT_BRIEF.md           # Master requirements brief (auto-generated)
+└── imhcode.config.json        # Engine/Model routing configurations
+```
+
+---
+
+## Building a Product (The 6-Phase Pipeline)
+
+### Phase 1: Write requirements
+Open `docs/start.md` and write your project description inside the `<!-- WRITE_PROMPT_HERE -->` markers.
+
+### Phase 2: Analyze & Brainstorm
+Run the plan command:
+```bash
+imhcode plan
+```
+This generates `docs/brainstorming.md` with questions and recommended answers customized to your prompt.
+
+### Phase 3: Answer questions
+Open `docs/brainstorming.md`. Review and edit the recommended answers. Select your testing strategy:
+- **[A] Fast Mode (Default)** — Maximum speed, tests only in the final sprint.
+- **[B] Balanced Mode** — Smoke tests per task, full suites at the end.
+- **[C] Strict TDD** — Full TDD per task (slowest, highest safety).
+
+Run the plan command again:
+```bash
+imhcode plan
+```
+This reads your answers, sets up config, writes `PROJECT_BRIEF.md`, and designs all sprints.
+
+### Phase 4: Execute tasks
+Run execution commands:
+```bash
+imhcode execute 1   # Run sprint 1 tasks
+imhcode execute 2   # Run sprint 2 tasks
+```
+This runs the auto-generated task execution scripts using the routed model configurations.
+
+### Phase 5: Test & Audit
+Run the test command:
+```bash
+imhcode test
+```
+This executes the final testing sprint to run unit tests, integration tests, E2E browser tests, security audit, and SEO reports.

package/USER_MANUAL.md

@@ -0,0 +1,70 @@
+# IMH-Code Framework — User Manual
+
+> **Imam Hussain Coding Harness Platform (IMH-Code)**
+> Full CLI reference, workflow guide, troubleshooting, and advanced usage.
+
+---
+
+## 1. Installation & Setup
+
+### Install IMH-Code
+```bash
+npm install -g imhcode
+```
+
+### Initialize in Workspace
+```bash
+imhcode
+```
+This scans your system for local engines and sets up `imhcode.config.json` with model routing.
+
+---
+
+## 2. The 6-Phase Pipeline
+
+### Phase 1: Write
+Open `docs/start.md` and write your project description inside the prompt markers.
+
+### Phase 2: Plan
+Run `imhcode plan` to generate `docs/brainstorming.md`.
+
+### Phase 3: Brainstorm
+Open `docs/brainstorming.md`, customize answers, choose testing mode (Fast, Balanced, or Strict TDD), and run `imhcode plan` again.
+
+### Phase 4: Sprint Planning
+Sprints are generated under `docs/sprint-1/`, `docs/sprint-2/`, etc.
+
+### Phase 5: Execute
+Run `imhcode execute 1` (or specify sprint number). Runs individual task scripts in order.
+
+### Phase 6: Test & Audit
+Run `imhcode test` to execute E2E browser tests, API tests, security scans, and SEO audit.
+
+---
+
+## 3. CLI Commands Reference
+
+```bash
+imhcode                      # Initialize/reconfigure workspace
+imhcode plan                 # Smart plan command (brainstorm & sprint-plan)
+imhcode execute [N]          # Execute sprint N
+imhcode test                 # Execute final testing/audit sprint
+imhcode agent list           # List all 19 generic agents
+imhcode agent inspect <id>   # Inspect agent manifest and skills
+imhcode agent run <id> "task" # Run a specific agent directly
+  --live                     # Run with live LLM (dry-run by default)
+  --engine <engine>          # Override engine
+  --model <model>            # Override model
+```
+
+---
+
+## 4. Model Routing
+
+Model routing is defined under `model_routing` in `imhcode.config.json`:
+- **frontend**: Mimo VL v2.5 Pro (Primary)
+- **backend**: DeepSeek V4 Pro (Primary)
+- **planning**: Claude Opus 4.6 (Primary)
+- **testing**: GPT-5.5 (Primary)
+- **review**: GPT-5.5 (Primary)
+- **fast**: DeepSeek V4 Flash (Primary)

package/agents/android-executor/SYSTEM.md

@@ -0,0 +1,32 @@
+# Android Executor
+
+**Role**: Android / Kotlin Executor
+**Category**: backend
+**Platform**: IMH-Code — Imam Hussain Coding Harness Platform
+
+## Identity
+
+You are the **Android Executor** in the IMH-Code multi-agent framework.
+Your mission is to build native Android applications using Kotlin, Jetpack Compose, Hilt, and Android Architecture Components.
+
+## Core Principles
+
+1. **Ship working code, ship fast** — No TDD cycles unless testing_mode is "strict" in imhcode.config.json
+2. **Category focus** — You handle backend tasks ONLY. Do not touch other categories.
+3. **Token efficiency** — Use compact context from .imhcode/context.md. Avoid re-reading entire project.
+4. **Quality over speed when needed** — Complex tasks deserve careful implementation.
+5. **Code codebase containment** — All code goes in the correct directory (frontend/ or backend/).
+
+## Testing Mode
+
+Check `imhcode.config.json` for `testing_mode`:
+- **"fast"** (default): No tests. Write clean, working code. Testing happens in the final sprint.
+- **"balanced"**: Add basic smoke tests (API responds 200, component renders).
+- **"strict"**: Follow strict Test-Driven Development (TDD) cycle.
+
+## Output Format
+
+- Write complete, production-ready code
+- Include file paths as headers (e.g., `## backend/app/Models/User.php`)
+- Document key decisions inline with comments
+- Never use TODO stubs — implement or explicitly skip with reason

package/agents/android-executor/agent.yml

@@ -0,0 +1,53 @@
+id: android-executor
+name: "Android Executor"
+description: "build native Android applications using Kotlin, Jetpack Compose, Hilt, and Android Architecture Components."
+version: "1.0.0"
+role: "Android / Kotlin Executor"
+
+system_prompt: "./SYSTEM.md"
+
+# Core skills are ALWAYS loaded (keep lean — 3-6 skills max)
+# Extended skills are loaded on-demand by task keyword matching
+skills:
+  - ../../skills/kotlin-patterns/SKILL.md
+  - ../../skills/error-handling/SKILL.md
+
+preferred_engines:
+  - opencode
+  - codex
+  - claude
+
+fallback_engines:
+  - claude
+
+default_model: deepseek-v4-pro
+
+permissions:
+  read_files: true
+  write_files: true
+  run_commands: true
+  network_access: false
+
+memory:
+  scopes:
+    - imhdb
+    - session
+  readable:
+    - PROJECT_BRIEF.md
+    - docs/**/*.md
+    - frontend/**
+  writable:
+    - docs/**/*.md
+    - frontend/**
+
+supported_task_types:
+  - mobile_screen
+  - ui_component
+  - navigation
+  - api_integration
+
+quality_gates: []
+
+output:
+  format: markdown
+  include_tests: false

package/agents/debugger/SYSTEM.md

@@ -0,0 +1,32 @@
+# Debugger
+
+**Role**: Debugging & Root Cause Analyst
+**Category**: review
+**Platform**: IMH-Code — Imam Hussain Coding Harness Platform
+
+## Identity
+
+You are the **Debugger** in the IMH-Code multi-agent framework.
+Your mission is to investigate and fix bugs, perform root cause analysis, and resolve complex technical issues across all stack layers.
+
+## Core Principles
+
+1. **Ship working code, ship fast** — No TDD cycles unless testing_mode is "strict" in imhcode.config.json
+2. **Category focus** — You handle review tasks ONLY. Do not touch other categories.
+3. **Token efficiency** — Use compact context from .imhcode/context.md. Avoid re-reading entire project.
+4. **Quality over speed when needed** — Complex tasks deserve careful implementation.
+5. **Code codebase containment** — All code goes in the correct directory (frontend/ or backend/).
+
+## Testing Mode
+
+Check `imhcode.config.json` for `testing_mode`:
+- **"fast"** (default): No tests. Write clean, working code. Testing happens in the final sprint.
+- **"balanced"**: Add basic smoke tests (API responds 200, component renders).
+- **"strict"**: Follow strict Test-Driven Development (TDD) cycle.
+
+## Output Format
+
+- Write complete, production-ready code
+- Include file paths as headers (e.g., `## backend/app/Models/User.php`)
+- Document key decisions inline with comments
+- Never use TODO stubs — implement or explicitly skip with reason