opencode-orchestrator 1.0.11 → 1.0.13

package/README.md

@@ -20,7 +20,7 @@ npm install -g opencode-orchestrator
 
 Then in OpenCode, launch your mission:
 ```bash
-/swarm "Architect and Build a Diablo-like Web Game"
+/task "Architect and Build a Diablo-like Web Game"
 ```
 
 ### 🌌 **The End of Sequential Limits. The Era of Fractal Mastery.**
@@ -88,14 +88,22 @@ MSVP ensures continuous integration integrity by decoupling verification from th
 *   **Global Barrier-Sync Protocol**: Implements a synchronized verification gate. The system enforces a blocking "Barrier" that requires consistency across all distributed units before transitioning to the final Master Review (E2E Integration) and Mission Sealing.
 
 ```
-              /swarm "Build Complex System"
+              /task "Build Diablo Game"
                            │
            ╔═══════════════════════════════════════╗
            ║     🎯 COMMANDER — Orchestrator       ║
            ║        (Main Session - Single)        ║
            ╚═══════════════════╤═══════════════════╝
                                │
-           ┌───────────────────▼───────────────────┐
+    ══════════════════════════▼══════════════════════════
+    ║          🚀 PHASE 0: DISCOVERY SWARM (Parallel)      ║
+    ══════════════════════════════════════════════════════
+    │               │               │               │
+    ▼ Scout         ▼ Scout         ▼ Scout         ▼ Scout
+ [Structure]      [Stack]         [Docs]          [Infra]
+    └───────────────┬───────────────┬───────────────┘
+                    │ Consolidate & Sync
+           ┌────────▼───────────────────┐
            │      📋 PLANNER — Create Plan         │
            │    → Outputs: Architectural Grid      │
            └───────────────────┬───────────────────┘

package/dist/index.js

@@ -631,7 +631,7 @@ var PROMPTS = {
 
 // src/shared/message/constants/slash-commands.ts
 var COMMAND_NAMES = {
-  TASK: "swarm",
+  TASK: "task",
   PLAN: "plan",
   STATUS: "status",
   STOP: "stop",
@@ -13312,6 +13312,12 @@ tool.schema = external_exports;
 var ENVIRONMENT_DISCOVERY = `${PROMPT_TAGS.ENVIRONMENT_DISCOVERY.open}
 **MANDATORY FIRST STEP** - Before any planning or coding:
 
+## \u26A1 Parallel Discovery Strategy
+To maximize cognitive velocity, execute Phase 0 as a **Parallel Discovery Swarm**:
+1. **Delegate** independent scouts (background=true) for Structure, Stack, and Infra discovery.
+2. **Consolidate** all empirical evidence once scouts complete their missions.
+3. **Verify** findings against existing codebase before sealing the initial context.
+
 ## 1. Project Structure Discovery
 Explore the project root to understand its organization:
 \`\`\`bash
@@ -13853,41 +13859,40 @@ ${PROMPT_TAGS.TOOLS.close}`;
 
 // src/agents/prompts/commander/execution.ts
 var COMMANDER_EXECUTION = `${PROMPT_TAGS.EXECUTION_STRATEGY.open}
-## Phase 0: ENVIRONMENT DISCOVERY (Never skip!)
-1. Initialize ${PATHS.OPENCODE}/ folder:
-   - If exists \u2192 DELETE and CREATE fresh
-   - If not exists \u2192 CREATE new
-   - This folder is for CONTEXT SHARING between agents
-2. Analyze project structure (ls, find)
-3. Read README.md, package.json, Dockerfile
-4. Identify build/test commands
-5. Save context to ${PATHS.CONTEXT}
-
-## Phase 1: THINK (Mandatory - Never Skip!)
-**THINK FIRST**: As ${AGENT_NAMES.COMMANDER}, think about ORCHESTRATION before action.
-
-### 1.1 MISSION SCOPE
-- What is the FULL scope of this mission?
-- What are the deliverables and success criteria?
-- What does the user REALLY want (not just what they said)?
-
-### 1.2 DECOMPOSITION
-- How can I break this into INDEPENDENT sub-tasks?
-- Which tasks MUST be sequential (dependencies)?
-- What is the MAXIMUM parallelism I can achieve?
-
-### 1.3 DELEGATION
-- Which agent is BEST for each task? (${AGENT_NAMES.PLANNER}/${AGENT_NAMES.WORKER}/${AGENT_NAMES.REVIEWER})
-- What context does each agent NEED to succeed?
-- What could cause an agent to FAIL or get stuck?
+## Phase 0: STRATEGIC SCOUT SWARM (Parallel Intelligence)
+**Launch Intelligence Scouts** to map the terrain before any strategic decisions.
+
+1. **Initialize \`${PATHS.OPENCODE}\`/ folder**: Clean start for synchronized context.
+2. **LAUNCH Scouts (background=true)**:
+   - **[Scout: Structure]**: Map directory tree and file patterns.
+   - **[Scout: Stack]**: Detect tech stack, build system, and dependencies.
+   - **[Scout: Docs]**: Analyze README, API docs, and architectural conventions.
+   - **[Scout: Infra]**: Identify CI/CD, Docker, and environment variables.
+3. **RECOVER & RETRIEVE**: Monitor scouts. Once finished, **READ \`${PATHS.CONTEXT}\` and \`ls -R\` outputs** to ingest all findings.
+
+## Phase 1: COMMANDER'S INTEL ANALYSIS (Strategic Mandate)
+**THINK FIRST**: Do not just blindly delegate. Analyze the gathered intelligence to set the mission trajectory.
+
+### 1.1 ARCHITECTURAL ANALYSIS
+- Based on Phase 0 reports, what is the existing architecture?
+- What are the core constraints and patterns I MUST enforce?
+- Identify potential bottlenecks or high-risk areas.
+
+### 1.2 THE STRATEGIC MANDATE
+- Define the **MISSION CHARTER** for the ${AGENT_NAMES.PLANNER}.
+- What is the specific architectural goal?
+- What are the non-negotiable success criteria?
+
+### 1.3 DECOMPOSITION
+- Break the mission into INDEPENDENT parallel streams.
+- Define the maximum concurrency slots for this specific mission.
 
 ### 1.4 RISK ASSESSMENT
 - What are the HIGH-RISK parts of this mission?
 - What is my FALLBACK if a task fails?
-- How will I DETECT and RECOVER from issues?
 - If agent fails \u2192 See ${PROMPT_TAGS.RECOVERY.open} section: DECOMPOSE and retry
 
-**ANTI-PATTERNS**: Sequential execution when parallel is possible. Doing work yourself instead of delegating. Starting without clear decomposition.
+**ANTI-PATTERNS**: Starting without intellectual synthesis. Sequential execution when parallel is possible.
 
 ## Phase 2: TRIAGE
 | Type | Signal | Approach |
@@ -13906,10 +13911,11 @@ ${AGENT_NAMES.PLANNER} creates ${PATHS.TODO} with parallel groups
 4. CONTINUE with dependent tasks
 5. REPEAT until all [ ] become [x]
 
-## Phase 5: VERIFY
-${AGENT_NAMES.REVIEWER} validates ALL work
-${AGENT_NAMES.REVIEWER} runs tests, confirms pass
-Only proceed to seal if PASS
+## Phase 5: STAGE 2 INTEGRATION REVIEW (Final Gate)
+1. **SYNC BARRIER**: Ensure all [Work + Stage 1 Review] pairs in ${PATHS.TODO} are SUCCESS.
+2. **VERIFY INTEGRATION**: ${AGENT_NAMES.REVIEWER} validates cross-module consistency.
+3. **E2E VALIDATION**: Run full system tests and build pass check.
+4. Only proceed to SEAL if all criteria are met with zero regressions.
 
 ## Phase 6: SEAL
 When ALL conditions met, output ${MISSION_SEAL.PATTERN}
@@ -14611,14 +14617,15 @@ ADAPTIVE IMPLEMENTATION WORKFLOW
 8. Handle edge cases like existing code does
 9. Add tests matching project's test style
 
-## Phase 5: VERIFY (Before reporting)
+## Phase 5: VERIFY & TRIGGER SHADOW REVIEW
 10. Run ${TOOL_NAMES.LSP_DIAGNOSTICS} \u2192 Must be clean
+11. Run BUILD & TEST commands from ${PATHS.CONTEXT}
+12. **TRIGGER STAGE 1 REVIEW**: 
+    - Before finishing, ${TOOL_NAMES.DELEGATE_TASK} to **${AGENT_NAMES.REVIEWER}** for unit verification.
+    - Prompt: "Verify unit changes in [your module] for mission [mission_id]. Run tests and check diagnostics."
+13. Report completion WITH evidence (test logs, review status).
 
-11. Run BUILD command from ${PATHS.CONTEXT}
-12. Run TEST command from ${PATHS.CONTEXT}
-13. Report completion WITH evidence
-
-**CRITICAL**: Do NOT mark [x] in ${PATHS.TODO} - that's ${AGENT_NAMES.REVIEWER}'s job!
+**CRITICAL**: Do NOT mark [x] in ${PATHS.TODO} - that's ${AGENT_NAMES.REVIEWER}'s job after Stage 2!
 ${PROMPT_TAGS.WORKFLOW.close}`;
 
 // src/agents/prompts/worker/quality.ts
@@ -15532,7 +15539,7 @@ $ARGUMENTS
 </execution_rules>
 </mission>`;
 var COMMANDS = {
-  "swarm": {
+  "task": {
     description: "MISSION MODE - Execute task autonomously until complete",
     template: MISSION_MODE_TEMPLATE,
     argumentHint: '"mission goal"'
@@ -15593,7 +15600,7 @@ THINK \u2192 PLAN \u2192 DELEGATE \u2192 EXECUTE \u2192 VERIFY \u2192 COMPLETE
 
 ## Usage
 - Select **${AGENT_NAMES.COMMANDER}** and type your request
-- Or use \`/swarm "your mission"\` explicitly
+- Or use \`/task "your mission"\` explicitly
 - ${AGENT_NAMES.COMMANDER} automatically coordinates all agents`
   }
 };
@@ -20549,7 +20556,7 @@ ${claudeRules}`;
 function detectSlashCommand(text) {
   const match = text.trim().match(/^\/([a-zA-Z0-9_-]+)(?:\s+(.*))?$/);
   if (!match) return null;
-  return { command: match[1], args: match[2] || "" };
+  return { command: match[1].toLowerCase(), args: match[2] || "" };
 }
 
 // src/utils/formatting/timestamp.ts
@@ -20587,7 +20594,7 @@ function createChatMessageHandler(ctx) {
     if (sessionID) {
       handleUserMessage(sessionID);
     }
-    if (agentName === AGENT_NAMES.COMMANDER) {
+    if (agentName === AGENT_NAMES.COMMANDER.toLowerCase()) {
       if (!sessions.has(sessionID)) {
         const now = Date.now();
         sessions.set(sessionID, {
@@ -20621,7 +20628,7 @@ function createChatMessageHandler(ctx) {
     }
     if (parsed) {
       const command = COMMANDS[parsed.command];
-      if (command && agentName !== AGENT_NAMES.COMMANDER) {
+      if (command && agentName !== AGENT_NAMES.COMMANDER.toLowerCase()) {
         parts[textPartIndex].text = command.template.replace(
           /\$ARGUMENTS/g,
           parsed.args || PROMPTS.CONTINUE

package/dist/shared/message/constants/slash-commands.d.ts

@@ -2,7 +2,7 @@
  * Slash Commands (with slash prefix)
  */
 export declare const SLASH_COMMANDS: {
-    readonly TASK: "/swarm";
+    readonly TASK: "/task";
     readonly PLAN: "/plan";
     readonly STATUS: "/status";
     readonly STOP: "/stop";
@@ -12,7 +12,7 @@ export declare const SLASH_COMMANDS: {
  * Command Names (without slash prefix, for comparison after parsing)
  */
 export declare const COMMAND_NAMES: {
-    readonly TASK: "swarm";
+    readonly TASK: "task";
     readonly PLAN: "plan";
     readonly STATUS: "status";
     readonly STOP: "stop";

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": "1.0.11",
+  "version": "1.0.13",
   "author": "agnusdei1207",
   "license": "MIT",
   "repository": {