fix: resolve audit findings — 10 bug fixes across scripts and docs

CRITICAL fixes:
- amoa_confirm_replacement.py: args.skip_ecos_notify → args.skip_amcos_notify, args.ecos_session → args.amcos_session
- Removed dead EXEC_STATE_FILE constant

HIGH fixes:
- stop_check/main.py: escalation print redirected to stderr (was corrupting JSON stdout)
- stop_check/main.py: removed dead CRITICAL_ERROR_OCCURRED variable
- amoa_reassign_kanban_tasks.py: implemented check_issue_has_open_pr using gh CLI, removed dead args

MEDIUM fixes:
- amoa_check_polling_due.py: datetime.now() → datetime.now(timezone.utc)
- amoa_sync_kanban.py: modules key lookup now tries modules_status first
- amoa_reassign_module.py: invalid message type "notification" → "info"
- amoa_orchestrator_stop_check.py: added "verified" and "skipped" to completion statuses
- stop_check/tasks.py: replaced naive regex with balanced-bracket parser

DOC fixes:
- ROLE_BOUNDARIES.md: AMCOS → AI Maestro/AMOA naming, bumped to 1.1.0
- AGENT_OPERATIONS.md: added amoa-kanban-management to skill table, bumped to 1.1.0

Version: 1.4.1

Author: Emasoft

.DS_Store

@@ -1,6 +1,6 @@
 {
   "name": "ai-maestro-orchestrator-agent",
-  "version": "1.4.0",
+  "version": "1.4.1",
   "description": "Task distribution, agent coordination, progress monitoring - executes plans via subagents. Requires AI Maestro for inter-agent messaging.",
   "author": {
     "name": "Emasoft",

.claude-plugin/plugin.json

@@ -1,124 +1,153 @@
-# Orchestrator Agent (amoa-)
+# AI Maestro Orchestrator Agent (amoa-)
 
-**Version**: 1.0.0
+**Version**: 1.4.0
 
 ## Overview
 
-The Orchestrator Agent handles **task distribution, agent coordination, and progress monitoring**. It receives plans from the Architect and coordinates subagents to implement them.
+The Orchestrator Agent handles **task distribution, agent coordination, and progress monitoring** for multi-agent projects. It receives plans and coordinates subagents to implement them via AI Maestro inter-agent messaging.
+
+## Requirements
+
+- Python 3.8+ with PyYAML (`pip install pyyaml`)
+- GitHub CLI (`gh`) for issue and project management
+- AI Maestro messaging system for inter-agent communication
 
 ## Core Responsibilities
 
-1. **Task Distribution**: Break plans into assignable tasks
-2. **Agent Coordination**: Manage subagents and remote agents
-3. **Progress Monitoring**: Track task completion
-4. **Module Management**: Organize work into modules
-5. **Verification**: Ensure instructions are followed correctly
+1. **Task Distribution**: Break plans into assignable tasks with dependency ordering
+2. **Agent Coordination**: Manage subagents and remote agents via AI Maestro
+3. **Progress Monitoring**: Track task completion via polling hooks
+4. **Module Management**: Organize work into GitHub Issue-backed modules
+5. **Verification**: Ensure instructions are followed correctly before exit
+6. **Kanban Management**: GitHub Projects V2 board and column management
 
 ## Components
 
-### Agents
+### Agents (6)
 
 | Agent | Description |
 |-------|-------------|
-| `amoa-main.md` | Main orchestrator agent |
-| `amoa-team-orchestrator.md` | Coordinates team of agents |
-| `amoa-task-summarizer.md` | Summarizes task progress |
-| `amoa-checklist-compiler.md` | Creates verification checklists |
-| `amoa-docker-container-expert.md` | Docker and container expertise |
-| `amoa-experimenter.md` | Experimentation and prototyping |
+| `amoa-orchestrator-main-agent` | Main orchestrator — delegates to sub-agents |
+| `amoa-team-orchestrator` | Coordinates team of remote agents |
+| `amoa-task-summarizer` | Summarizes task progress and state |
+| `amoa-checklist-compiler` | Creates verification checklists |
+| `amoa-docker-container-expert` | Docker and container expertise |
+| `amoa-experimenter` | Experimentation and prototyping |
 
-### Commands
+### Commands (15)
 
 | Command | Description |
 |---------|-------------|
-| `amoa-start-orchestration` | Start orchestration phase |
-| `amoa-register-agent` | Register remote agent |
-| `amoa-assign-module` | Assign module to agent |
-| `amoa-reassign-module` | Reassign module |
-| `amoa-check-agents` | Check agent status |
-| `amoa-add-module` | Add new module |
-| `amoa-modify-module` | Modify module |
-| `amoa-remove-module` | Remove module |
-| `amoa-prioritize-module` | Change module priority |
-| `amoa-orchestrator-loop` | Start orchestration loop |
-| `amoa-orchestrator-status` | Check orchestrator status |
-| `amoa-cancel-orchestrator` | Cancel orchestration |
-
-### Skills
+| `/amoa-start-orchestration` | Start orchestration phase |
+| `/amoa-orchestrator-loop` | Start orchestration loop |
+| `/amoa-orchestration-status` | Show orchestration phase status (modules/agents) |
+| `/amoa-orchestrator-status` | Show orchestrator loop status (iterations/tasks) |
+| `/amoa-cancel-orchestrator` | Cancel orchestration |
+| `/amoa-register-agent` | Register remote agent |
+| `/amoa-check-agents` | Check agent status and poll progress |
+| `/amoa-assign-module` | Assign module to agent |
+| `/amoa-reassign-module` | Reassign module to different agent |
+| `/amoa-add-module` | Add new module |
+| `/amoa-modify-module` | Modify existing module |
+| `/amoa-remove-module` | Remove module |
+| `/amoa-prioritize-module` | Change module priority |
+| `/amoa-reassign-kanban-tasks` | Reassign kanban board tasks |
+| `/amoa-generate-replacement-handoff` | Generate agent replacement handoff document |
+
+### Skills (16)
 
 | Skill | Description |
 |-------|-------------|
-| `amoa-two-phase-mode` | Two-phase orchestration mode |
-| `amoa-orchestration-commands` | Orchestration command patterns |
-| `amoa-orchestration-patterns` | Orchestration best practices |
-| `amoa-remote-agent-coordinator` | Remote agent coordination |
-| `amoa-module-management` | Module CRUD operations |
-| `amoa-verification-patterns` | Instruction verification |
-| `amoa-developer-communication` | Developer comm patterns |
-| `amoa-checklist-compilation-patterns` | Checklist generation |
-| `amoa-agent-replacement` | Agent replacement handoffs |
-| `amoa-task-distribution` | Task breakdown and assignment |
-| `amoa-progress-monitoring` | Progress tracking and polling |
-| `amoa-messaging-templates` | AI Maestro message formats |
-| `amoa-label-taxonomy` | GitHub label system |
-| `amoa-implementer-interview-protocol` | Interview-based requirements |
-| `amoa-github-action-integration` | GitHub Actions patterns |
-
-### Hooks
+| `amoa-two-phase-mode` | Plan-then-Execute workflows with formal approval |
+| `amoa-orchestration-commands` | Orchestration command reference and loop mechanics |
+| `amoa-orchestration-patterns` | Task breakdown patterns for human developers |
+| `amoa-remote-agent-coordinator` | Remote AI agent coordination via AI Maestro |
+| `amoa-module-management` | Module CRUD operations (1:1 with GitHub Issues) |
+| `amoa-verification-patterns` | Implementation verification and evidence collection |
+| `amoa-developer-communication` | Developer communication patterns and templates |
+| `amoa-checklist-compilation-patterns` | Verification checklist generation |
+| `amoa-agent-replacement` | Agent replacement and handoff protocols |
+| `amoa-task-distribution` | Task breakdown, assignment, and load balancing |
+| `amoa-progress-monitoring` | Progress tracking via state-based polling |
+| `amoa-messaging-templates` | AI Maestro message format templates |
+| `amoa-label-taxonomy` | GitHub label system for multi-agent coordination |
+| `amoa-implementer-interview-protocol` | Interview-based task verification |
+| `amoa-github-action-integration` | GitHub Actions CI/CD patterns |
+| `amoa-kanban-management` | GitHub Projects V2 kanban board management |
+
+### Hooks (4)
 
 | Hook | Event | Description |
 |------|-------|-------------|
-| `amoa-orchestrator-stop` | Stop | Block exit until tasks complete |
-| `amoa-instruction-verification-check` | PreToolUse | Verify instructions before work |
-| `amoa-polling-reminder` | UserPromptSubmit | Remind to poll progress |
-| `amoa-file-tracker` | PostToolUse | Track file modifications |
-
-## Project Structure Notes
-
-### Non-Standard Directories
-
-- **`git-hooks/`** — Contains git hook scripts (pre-push) for plugin validation before pushing
-- **`shared/`** — Shared resources used across skills and agents
+| `amoa-orchestrator-stop` | Stop | Blocks exit until all tasks complete (120s timeout) |
+| `amoa-instruction-verification-check` | PreToolUse | Blocks agent work if verification incomplete |
+| `amoa-polling-reminder` | UserPromptSubmit | Reminds to poll agent progress |
+| `amoa-file-tracker` | PostToolUse | Tracks Edit/Write file modifications |
 
-### Shell Scripts Compatibility
+## Project Structure
 
-3 bash/shell scripts (`.sh` files) require **Linux or macOS** and are not natively available on Windows. Windows users should use WSL2 or equivalent.
+```
+ai-maestro-orchestrator-agent/
+├── .claude-plugin/      # Plugin manifest (plugin.json)
+├── agents/              # Agent definitions (6 agents)
+├── commands/            # Slash command definitions (15 commands)
+├── docs/                # Architecture docs (role boundaries, workflow, registry)
+├── git-hooks/           # Git hooks (pre-push validation)
+├── hooks/               # Plugin hooks (hooks.json + 4 hook scripts)
+├── scripts/             # Python scripts (37 scripts + validation suite)
+│   └── amoa_stop_check/ # Stop hook package (lock, tasks, phase checks)
+├── shared/              # Shared resources across skills
+├── skills/              # Skill definitions (16 skills with references)
+├── tests/               # Test suite
+└── requirements.txt     # Python dependencies (PyYAML)
+```
 
 ## Workflow
 
-1. Receives plan from Architect (via Assistant Manager)
-2. Creates modules from plan
-3. Assigns modules to agents
-4. Monitors progress via polling
-5. Handles failures and reassignments
-6. Reports completion to Assistant Manager
-7. Hands off to Integrator for quality gates
+1. **Plan Phase**: Define modules and requirements (two-phase mode)
+2. **Assignment**: Assign modules to agents via kanban board
+3. **Execution**: Agents implement modules, report via AI Maestro
+4. **Monitoring**: Poll agent progress, detect stalls, handle failures
+5. **Verification**: 4-loop verification ensures all tasks complete
+6. **Completion**: Stop hook enforces completion before session exit
 
-## Installation (Production)
+## Installation
 
-<!-- Marketplace installation will be available once the AI Maestro marketplace is configured -->
+Install using `--plugin-dir` for local development:
 
-Install using `--plugin-dir` for now (marketplace coming soon). Use `--scope local` to install only for this agent's directory, or `--scope global` for all projects.
-
-Role plugins are installed with `--scope local` inside the specific agent's working directory (`~/agents/<agent-name>/`). This ensures the plugin is only available to that agent.
+```bash
+claude --plugin-dir /path/to/ai-maestro-orchestrator-agent
+```
 
-Once installed, start a session with the main agent:
+Start a session with the main orchestrator agent:
 
 ```bash
 claude --agent amoa-orchestrator-main-agent
 ```
 
-## Development Only (--plugin-dir)
-
-`--plugin-dir` loads a plugin directly from a local directory without marketplace installation. Use only during plugin development.
+After modifying plugin files, reload without restarting:
 
 ```bash
-claude --plugin-dir ./OUTPUT_SKILLS/ai-maestro-orchestrator-agent
+/reload-plugins
 ```
 
 ## Validation
 
+Run the CPV (Claude Plugins Validation) suite:
+
 ```bash
-cd OUTPUT_SKILLS/ai-maestro-orchestrator-agent
-uv run python scripts/validate_plugin.py . --verbose
+cd ai-maestro-orchestrator-agent
+uv run --with pyyaml --with mypy --with types-PyYAML python scripts/validate_plugin.py . --verbose --strict
 ```
+
+The pre-push hook automatically syncs validators and runs validation before each push.
+
+## Development
+
+- **Sync validators**: `python3 scripts/sync_cpv_validators.py`
+- **Kanban management**: `python3 scripts/amoa_kanban_manager.py <command>`
+- **GitHub project columns**: `python3 scripts/gh-project-add-columns.py`
+
+## License
+
+MIT

README.md

@@ -186,6 +186,7 @@ Claude Code's skill resolution system automatically finds skills by folder name
 | `amoa-remote-agent-coordinator` | Coordinating remote AI agents |
 | `amoa-task-distribution` | Distributing tasks to implementer agents |
 | `amoa-two-phase-mode` | Two-phase planning and execution mode |
+| `amoa-kanban-management` | GitHub Projects V2 kanban board management (create boards, add columns, move items) |
 | `amoa-verification-patterns` | Code and deliverable verification patterns |
 
 ### NEVER Reference Other Plugins
@@ -458,6 +459,10 @@ The following skills were added to AMOA (2026-02-06 — 2026-02-07):
 
 ---
 
+## Recent Changes (2026-03-06)
+
+- Added `amoa-kanban-management` to Available AMOA Skills table, updated version to 1.1.0
+
 ## Recent Changes (2026-02-07)
 
 - Added 8-column canonical kanban system (unified from 5 conflicting systems)
@@ -487,6 +492,6 @@ The following skills were added to AMOA (2026-02-06 — 2026-02-07):
 
 ---
 
-**Document Version**: 1.0.0
-**Last Updated**: 2026-02-04
+**Document Version**: 1.1.0
+**Last Updated**: 2026-03-06
 **Maintained By**: claude-skills-factory

docs/AGENT_OPERATIONS.md

@@ -1,4 +1,4 @@
-# AMCOS Role Boundaries
+# AI Maestro Role Boundaries
 
 **CRITICAL: This document defines the strict boundaries between agent roles. Violating these boundaries breaks the system architecture.**
 
@@ -191,6 +191,6 @@ AMOA: Sends handoff to agent-456
 
 ---
 
-**Document Version**: 1.0.0
-**Last Updated**: 2026-02-02
-**Author**: AMCOS Plugin Development
+**Document Version**: 1.1.0
+**Last Updated**: 2026-03-06
+**Author**: AMOA Plugin Development

docs/ROLE_BOUNDARIES.md

@@ -16,7 +16,7 @@
 
 import json
 import sys
-from datetime import datetime, timedelta
+from datetime import datetime, timedelta, timezone
 from pathlib import Path
 
 # State file path (relative to project root)
@@ -99,7 +99,7 @@ def check_polling_status() -> tuple[list, list]:
     if not active_assignments:
         return [], []
 
-    now = datetime.now()
+    now = datetime.now(timezone.utc)
     overdue = []
     warning = []
 

scripts/amoa_check_polling_due.py

@@ -58,8 +58,6 @@
 # State file location relative to the project root (JSON format)
 STATE_FILE_PATH = ".ai-maestro/orchestration-state.json"
 
-# Fallback exec-phase state file (YAML frontmatter in markdown)
-EXEC_STATE_FILE = Path(".claude/orchestrator-exec-phase.local.md")
 
 # Audit log location relative to the project root
 AUDIT_LOG_DIR = "logs"
@@ -616,7 +614,7 @@ def main():
 
     # Step 3: Send AMCOS notification
     ecos_notified = False
-    if not args.skip_ecos_notify:
+    if not args.skip_amcos_notify:
         notification_details = {
             "ack_status": ack_status,
             "state_updated": state is not None,
@@ -627,7 +625,7 @@ def main():
             notification_details["assignments_updated"] = state_updates.get("assignments_updated", 0)
 
         ecos_notified = send_ecos_notification(
-            ecos_session=args.ecos_session,
+            ecos_session=args.amcos_session,
             status=outcome,
             failed_agent=args.failed_agent,
             new_agent=args.new_agent,

scripts/amoa_confirm_replacement.py

@@ -140,7 +140,7 @@ def check_orchestration_phase(state):
     for module in modules:
         module_id = module.get("id", "unknown")
         module_status = module.get("status", "unknown")
-        if module_status not in ("complete", "done"):
+        if module_status not in ("complete", "done", "verified", "skipped"):
             incomplete_modules.append({"id": module_id, "status": module_status})
 
     if incomplete_modules: