opencode-orchestrator 0.5.6 → 0.5.8

package/README.md

@@ -11,34 +11,18 @@
 
 ---
 
-## Why I Built This 🤔
+## Why?
 
-**I was frustrated coding with DeepSeek and Z.AI.**
-
-I wanted to achieve the same quality of work as premium models like **Gemini 3 Pro** and **Claude Opus**, but with affordable models.
-For developers who chose **budget-friendly subscriptions** instead of expensive plans, I built a multi-agent system that can **autonomously complete** complex engineering tasks even with mid-tier LLMs.
-
-> *"Intelligence is a resource. Orchestrate it."*
-
----
-
-## What is this?
-
-A **5-agent autonomous architecture** designed to solve complex engineering tasks with high reliability, even on mid-range LLMs.
-
-**Core Philosophy**: Intelligence is a resource. We orchestrate that resource through **Phase-based Workflows** and **Mandatory Environment Scans** to ensure code always fits the project's infrastructure.
-
-> 🦀 **Powered by Rust** — Background tasks and parallel searches run on native Rust binaries for maximum performance.
+Tested GLM-4, got disappointed. Built this to make mid-tier models work like premium ones through structured orchestration.
 
 ### Key Features
-- **🎯 Autonomous Loop** — Commander runs relentlessly until the mission is complete.
-- **🔍 Environment Scan** — Mandatory analysis of Infra (Docker/OS), Stack, and Domain before any code change.
-- **🔨 Smart Implementation** — Builder matches existing codebase patterns exactly.
-- **🛡️ Rigorous Audit** — Inspector proves success with environment-specific evidence (Builds/Tests/Logs).
-- **💾 Persistent Context** — Recorder saves session state to disk, enabling resume at any time.
-- **🏗️ Parallel Agents** — Delegated agent execution (`delegate_task`) with sync/async modes.
-- **⏳ Background Tasks** — Run long commands (builds, tests) in background and check results later.
-- **🔎 mgrep** — Multi-pattern parallel search powered by Rust for blazing-fast codebase analysis.
+- **🎯 Autonomous Loop** — Commander runs until the mission is complete
+- **🔍 Environment Scan** — Analyzes Infra, Stack, and Domain before coding
+- **🔨 Smart Implementation** — Matches existing codebase patterns
+- **🛡️ Rigorous Audit** — Proves success with builds/tests/logs
+- **💾 Persistent Context** — Saves session state to disk
+- **🏗️ Parallel Agents** — Run multiple agents concurrently
+- **⏳ Background Tasks** — Non-blocking command execution
 
 ---
 
@@ -54,168 +38,33 @@ Restart OpenCode after installation.
 
 ## Usage
 
-### 🚀 Method 1: Select Commander via Tab Key (Recommended)
+### 🚀 Select Commander via Tab Key (Recommended)
 
-In OpenCode, press `Tab` to open the Agent selection menu. Select **Commander** and type your mission!
-
-<div align="center">
-  <img src="assets/commander-screenshot.png" alt="Commander Screenshot" width="600" />
-  <p><em>Press Tab to select Commander</em></p>
-</div>
+Press `Tab` in OpenCode → Select **Commander** → Type your mission!
 
 ```
 "Fix the login bug in the docker-compose environment"
 ```
 
-The Commander will:
-1. **Survey**: Scan the Docker environment and codebase.
-2. **Plan**: Break the fix into steps.
-3. **Execute**: Call Builder to fix while matching patterns.
-4. **Verify**: Run builds/tests to prove the fix works.
-5. **Complete**: Report results with concrete evidence.
-
-### 📋 Method 2: Use /task Command
+### 📋 Use /task Command
 
 ```bash
 /task "Implement user authentication with JWT"
 ```
 
-> **💡 Tip:** Using the `/task` command makes Commander mode run **2x longer**.
-> Use `/task` for complex tasks that need extended processing!
-
-### ⚡ Example: Parallel Execution
-
-Trigger parallel agent execution with prompts like:
-
-```
-"Build and test in parallel"
-"Implement feature X while reviewing module Y"
-"Run linting, tests, and build at the same time"
-```
-
-Commander will automatically use `delegate_task` with `background: true` for independent tasks.
-
-Monitor parallel tasks in the terminal:
-```
-[parallel] 🚀 SPAWNED task_a1b2 → builder: Implement feature X
-[parallel] 🚀 SPAWNED task_c3d4 → inspector: Review module Y
-[parallel] ✅ COMPLETED task_c3d4 → inspector: Review module Y (45s)
-[parallel] 🗑️ CLEANED task_c3d4 (session deleted)
-```
-
----
-
-## Agents (5-Agent Architecture)
-
-| Agent | Emoji | Role | Responsibility |
-| :--- | :--- | :--- | :--- |
-| **Commander** | 🎯 | Orchestrator | Autonomous mission control & delegation |
-| **Architect** | 🏗️ | Planner | Task decomposition & strategy correction |
-| **Builder** | 🔨 | Developer | Full-stack implementation (Logic + UI) |
-| **Inspector** | 🔍 | Quality | 5-point audit & automatic bug fixing |
-| **Recorder** | 💾 | Context | Persistent environment & progress tracking |
-
----
-
-## The Workflow (Progressive Phases)
-
-1. **Phase 0: Triage (Smart)**
-   - **Fast Track 🟢**: Simple fixes → Execute instantly (Skip heavy scans).
-   - **Deep Track 🔴**: Complex features → Full Environment Scan & Plan.
-2. **Phase 1: Environment Scan**
-   - Mandatory for Deep Track: Infra/Domain/Stack analysis.
-3. **Phase 2: Parallel Planning**
-   - Architect creates a DAG of atomic tasks (Scalable Planning).
-4. **Phase 3: Execution & Audit**
-   - Builder writes code ↔ Inspector verifies with evidence.
-5. **Phase 4: Completion**
-   - Mission Complete reported with proof of build/test success.
+> **💡 Tip:** `/task` makes Commander run **2x longer** for complex tasks.
 
 ---
 
-## Development
-
-### Quick Start (Local Development)
-
-```bash
-# Step 1: Build and link (run this ONCE)
-npm run dev:link
-
-# Step 2: Restart OpenCode
-
-# Step 3: Make code changes...
-
-# Step 4: Rebuild and re-link (after code changes)
-npm run dev:link
-
-# (Optional) Check link status
-npm run dev:status
-```
-
-### Development Scripts
-
-```bash
-# Build the plugin
-npm run build
-
-# Run all tests (87 tests)
-npm run test:all
-
-# Run unit tests only
-npm run test:unit
-
-# Run E2E tests only
-npm run test:e2e
-
-# Link for local testing (auto-rebuilds)
-npm run dev:link
-
-# Unlink from global
-npm run dev:unlink
-
-# Check if plugin is linked globally
-npm run dev:status
-```
-
-### Test Suite
-
-| Command | Description | Tests |
-|---------|-------------|-------|
-| `npm run test:all` | Run all tests with verbose output | 87 |
-| `npm run test:unit` | Unit tests (concurrency, task-store, etc.) | 40 |
-| `npm run test:e2e` | E2E tests (background, session, rust, system) | 47 |
-| `npm run test:coverage` | Generate coverage report | - |
-
-### Development Workflow
-
-| Step | Command | Description |
-|------|----------|-------------|
-| 1️⃣ Initial setup | `npm run dev:link` | Build + link for the first time |
-| 2️⃣ Test | Open OpenCode | Test your changes |
-| 3️⃣ Make changes | Edit code | Modify TypeScript files |
-| 4️⃣ Rebuild | `npm run dev:link` | Rebuild + re-link |
-| 5️⃣ Repeat | 2-4 | Iterate on changes |
-| ❌ Cleanup | `npm run dev:unlink` | Unlink when done |
-
-### Release Scripts
-
-```bash
-# Release patch version (0.5.4 → 0.5.5)
-npm run release:patch
-
-# Release minor version (0.5.5 → 0.6.0)
-npm run release:minor
-
-# Release major version (0.5.5 → 1.0.0)
-npm run release:major
-```
+## The 5 Agents
 
-Each release script:
-1. Builds the plugin
-2. Bumps version in package.json
-3. Creates git tag
-4. Pushes to remote
-5. Publishes to npm
+| Agent | Role | Responsibility |
+| :--- | :--- | :--- |
+| **Commander** 🎯 | Orchestrator | Autonomous mission control |
+| **Architect** 🏗️ | Planner | Task decomposition |
+| **Builder** 🔨 | Developer | Full-stack implementation |
+| **Inspector** 🔍 | Quality | Audit & auto-fix |
+| **Recorder** 💾 | Context | Progress tracking |
 
 ---
 
@@ -227,11 +76,10 @@ npm uninstall -g opencode-orchestrator
 
 ---
 
-## Docs
+## Documentation
 
-- [Architecture & Design](docs/ARCHITECTURE.md) — Detailed system design and agent protocols
-- [Plugin Troubleshooting](docs/PLUGIN_TROUBLESHOOTING.md) — Setup and common issues
-- [Changelog](CHANGELOG.md) — Version history and updates
+- [Architecture & Design](docs/ARCHITECTURE.md)
+- [Troubleshooting](docs/PLUGIN_TROUBLESHOOTING.md)
 
 ---
 

package/dist/index.js

@@ -22,11 +22,12 @@ You are Commander. Complete missions autonomously. Never stop until done.
 </role>
 
 <core_rules>
-1. Never stop until "\u2705 MISSION COMPLETE"
-2. Never wait for user during execution
-3. Never stop because agent returned nothing
-4. Always survey environment & codebase BEFORE coding
-5. Always verify with evidence based on runtime context
+1. LANGUAGE: ALL output MUST be in English only. No exceptions. No other languages.
+2. Never stop until "\u2705 MISSION COMPLETE"
+3. Never wait for user during execution
+4. Never stop because agent returned nothing
+5. Always survey environment & codebase BEFORE coding
+6. Always verify with evidence based on runtime context
 </core_rules>
 
 <phase_0 name="TRIAGE">
@@ -216,8 +217,8 @@ You are Architect. Break complex tasks into atomic pieces.
 </role>
 
 <constraints>
-
-If your reasoning collapses into gibberish, stop and output "ERROR: REASONING_COLLAPSE".
+1. LANGUAGE: ALL output MUST be in English only. No exceptions. No other languages.
+2. If your reasoning collapses into gibberish, stop and output "ERROR: REASONING_COLLAPSE".
 </constraints>
 
 <scalable_planning>
@@ -278,8 +279,8 @@ You are Builder. Write code that works.
 </role>
 
 <constraints>
-
-If your reasoning collapses into gibberish, stop and output "ERROR: REASONING_COLLAPSE".
+1. LANGUAGE: ALL output MUST be in English only. No exceptions. No other languages.
+2. If your reasoning collapses into gibberish, stop and output "ERROR: REASONING_COLLAPSE".
 </constraints>
 
 <scalable_attention>
@@ -344,8 +345,8 @@ You are Inspector. Prove failure or success with evidence.
 </role>
 
 <constraints>
-
-If your reasoning collapses into gibberish, stop and output "ERROR: REASONING_COLLAPSE".
+1. LANGUAGE: ALL output MUST be in English only. No exceptions. No other languages.
+2. If your reasoning collapses into gibberish, stop and output "ERROR: REASONING_COLLAPSE".
 </constraints>
 
 <scalable_audit>
@@ -411,8 +412,8 @@ You are Recorder. Save and load work progress.
 </role>
 
 <constraints>
-
-If your reasoning collapses into gibberish, stop and output "ERROR: REASONING_COLLAPSE".
+1. LANGUAGE: ALL output MUST be in English only. No exceptions. No other languages.
+2. If your reasoning collapses into gibberish, stop and output "ERROR: REASONING_COLLAPSE".
 </constraints>
 
 <purpose>
@@ -13053,6 +13054,10 @@ var COMMANDS = {
 You are Commander. Complete this mission. Never stop until 100% done.
 </role>
 
+<language_rule>
+CRITICAL: ALL output MUST be in English only. No exceptions. No other languages permitted.
+</language_rule>
+
 <phase_1 name="MANDATORY_ENVIRONMENT_SCAN">
 Before any planning or coding, you MUST understand:
 1. INFRA: OS-native? Container? Docker-compose? Volume-mounted?

package/package.json

@@ -2,7 +2,7 @@
   "name": "opencode-orchestrator",
   "displayName": "OpenCode Orchestrator",
   "description": "Distributed Cognitive Architecture for OpenCode. Turns simple prompts into specialized multi-agent workflows (Planner, Coder, Reviewer).",
-  "version": "0.5.6",
+  "version": "0.5.8",
   "author": "agnusdei1207",
   "license": "MIT",
   "repository": {