cpa-codex-manager-openai-pool
Manage OpenAI account pools with automated registration, CLIProxyAPI monitoring, and intelligent auto-replenishment
How do I install this agent skill?
npx skills add https://github.com/reason-machines/codex-skills --skill cpa-codex-manager-openai-poolIs this agent skill safe to install?
- Gen Agent Trust Hubpass
This skill provides a management panel for OpenAI account pools, focusing on automation, monitoring, and maintenance. It uses standard Python libraries for web services, database management, and asynchronous task handling.
- Socketwarn
1 alert: gptSecurity
- Snykwarn
Risk: MEDIUM · 1 issue
What does this agent skill do?
CPA-Codex-Manager OpenAI Account Pool Management
Skill by ara.so — Codex Skills collection.
CPA-Codex-Manager is a high-performance management panel for OpenAI account pools, featuring automated batch registration, real-time CLIProxyAPI platform monitoring, and intelligent maintenance systems.
What It Does
- Batch Account Registration: Parallel (up to 50 threads, 1000 tasks) and pipeline modes with random intervals
- CLIProxyAPI Auto-Inspection: Detects 401 auth failures and quota exhaustion, auto-removes invalid accounts
- Smart Auto-Replenishment: Monitors pool health, triggers automated replenishment when accounts fall below threshold
- Real-time Monitoring: WebSocket-based log streaming and progress visualization
- Multi-Email Support: Integrates Outlook, TempMail, CloudMail services
- Emergency Defense: Dynamic threshold protection with configurable cooling periods
Installation
Standard Installation
# Using uv (recommended)
uv sync
# Or using pip
pip install -r requirements.txt
Environment Configuration
cp .env.example .env
Edit .env with your configuration:
# Web UI Configuration
WEBUI_HOST=0.0.0.0
WEBUI_PORT=8000
WEBUI_ACCESS_PASSWORD=your_secure_password
# Database (SQLite default)
APP_DATABASE_URL=data/database.db
# Or PostgreSQL
# APP_DATABASE_URL=postgresql://user:password@host:5432/dbname
Docker Installation
mkdir -p ~/CPA-Codex-Manager/data ~/CPA-Codex-Manager/logs
cd ~/CPA-Codex-Manager
curl -O https://raw.githubusercontent.com/Maoleio/CPA-Codex-Manager/main/docker-compose.yml
docker compose up -d
Desktop Mode
pip install pywebview
python desktop.py
Running the Application
Web Mode
python webui.py
Access at http://localhost:8000
Docker Mode
docker compose up -d
docker compose logs -f # View logs
docker compose pull && docker compose up -d # Update
Core Python API Usage
Database Models
from app.models import (
AccountPool,
RegistrationTask,
CPAService,
InspectionHistory
)
from app.database import get_db
# Query account pool
with get_db() as db:
accounts = db.query(AccountPool).filter(
AccountPool.status == 'ready'
).all()
print(f"Ready accounts: {len(accounts)}")
for acc in accounts:
print(f"Email: {acc.email}, Token: {acc.refresh_token}")
Registration Task Creation
from app.services.registration import RegistrationService
from app.schemas import RegistrationTaskCreate
async def create_batch_registration():
service = RegistrationService()
task_data = RegistrationTaskCreate(
count=100,
email_service="outlook", # outlook, tempmail, cloudmail
mode="parallel", # parallel or pipeline
max_workers=50,
random_interval_min=5,
random_interval_max=15
)
task = await service.create_task(task_data)
print(f"Task created: {task.id}")
# Start registration
await service.start_task(task.id)
CPA Service Integration
from app.services.cpa_service import CPAServiceManager
async def configure_cpa():
manager = CPAServiceManager()
# Add CPA service
cpa = await manager.add_service(
name="Production CPA",
base_url="https://your-cpa-instance.com",
admin_token="${CPA_ADMIN_TOKEN}",
enabled=True
)
# Configure auto-inspection
await manager.update_inspection_config(
cpa.id,
enabled=True,
interval_minutes=60,
check_401=True,
check_quota=True,
auto_remove=True
)
Auto-Replenishment Configuration
from app.services.replenishment import ReplenishmentService
async def setup_auto_replenish():
service = ReplenishmentService()
# Configure replenishment rules
await service.configure(
cpa_service_id=1,
enabled=True,
threshold=50, # Trigger when ready accounts < 50
replenish_count=100,
mode="parallel",
email_service="outlook"
)
Manual Account Inspection
from app.services.inspection import InspectionService
async def run_inspection():
service = InspectionService()
result = await service.inspect_cpa_accounts(
cpa_service_id=1,
check_401=True,
check_quota=True,
auto_remove=True
)
print(f"Inspection complete:")
print(f" Total checked: {result['total_checked']}")
print(f" Valid: {result['valid_count']}")
print(f" Invalid (401): {result['invalid_401']}")
print(f" Quota exhausted: {result['quota_exhausted']}")
print(f" Removed: {result['removed_count']}")
Emergency Defense Configuration
from app.services.emergency import EmergencyDefenseService
async def configure_emergency_defense():
service = EmergencyDefenseService()
await service.configure(
enabled=True,
threshold_percentage=50, # Trigger if ready% < 50%
cleanup_percentage=50, # Remove 50% of accounts
cooldown_minutes=5 # Wait 5min before retry
)
WebSocket Integration for Real-time Logs
from fastapi import WebSocket
import asyncio
# Server-side WebSocket handler
async def websocket_log_stream(websocket: WebSocket, task_id: int):
await websocket.accept()
# Stream registration logs
async for log_line in get_task_logs(task_id):
await websocket.send_json({
"type": "log",
"task_id": task_id,
"message": log_line,
"timestamp": datetime.utcnow().isoformat()
})
await websocket.close()
Common Patterns
Full Registration Workflow
from app.services.registration import RegistrationService
from app.services.cpa_service import CPAServiceManager
async def full_registration_workflow():
# Create and start registration task
reg_service = RegistrationService()
task = await reg_service.create_task(
count=200,
email_service="outlook",
mode="parallel",
max_workers=50
)
# Monitor progress
await reg_service.start_task(task.id)
while True:
status = await reg_service.get_task_status(task.id)
print(f"Progress: {status['success']}/{status['total']} "
f"({status['progress_percentage']:.1f}%)")
if status['status'] in ['completed', 'failed']:
break
await asyncio.sleep(5)
# Upload successful accounts to CPA
cpa_manager = CPAServiceManager()
uploaded = await cpa_manager.upload_accounts(
cpa_service_id=1,
task_id=task.id
)
print(f"Uploaded {uploaded} accounts to CPA")
Scheduled Inspection with Auto-Replenish
import asyncio
from app.services.inspection import InspectionService
from app.services.replenishment import ReplenishmentService
async def scheduled_maintenance():
inspection = InspectionService()
replenish = ReplenishmentService()
while True:
# Run inspection
result = await inspection.inspect_cpa_accounts(
cpa_service_id=1,
check_401=True,
check_quota=True,
auto_remove=True
)
# Check if replenishment needed
ready_count = result['valid_count']
if ready_count < 50:
print(f"Low account count ({ready_count}), triggering replenishment")
await replenish.trigger_replenish(
cpa_service_id=1,
count=100
)
# Wait 1 hour
await asyncio.sleep(3600)
Query Account Pool Statistics
from app.database import get_db
from app.models import AccountPool
from sqlalchemy import func
def get_pool_statistics():
with get_db() as db:
stats = db.query(
AccountPool.status,
func.count(AccountPool.id).label('count')
).group_by(AccountPool.status).all()
return {status: count for status, count in stats}
# Example output: {'ready': 150, 'invalid': 10, 'quota_exhausted': 5}
Configuration Best Practices
Inspection Configuration
# config.py or .env
INSPECTION_CONFIG = {
"interval_minutes": 60, # Check every hour
"check_401": True, # Enable auth failure detection
"check_quota": True, # Enable quota exhaustion detection
"auto_remove": True, # Auto-remove invalid accounts
"emergency_threshold": 50, # Trigger emergency if ready% < 50%
}
Replenishment Configuration
REPLENISHMENT_CONFIG = {
"enabled": True,
"threshold": 50, # Trigger when ready accounts < 50
"replenish_count": 100, # Add 100 accounts per trigger
"mode": "parallel", # Use parallel mode for speed
"email_service": "outlook",
"max_workers": 50
}
Registration Modes
Parallel Mode (fast, high volume):
parallel_config = {
"mode": "parallel",
"max_workers": 50,
"count": 1000
}
Pipeline Mode (stealth, bypass rate limits):
pipeline_config = {
"mode": "pipeline",
"random_interval_min": 10,
"random_interval_max": 30,
"count": 100
}
Troubleshooting
Registration Failures
# Check failed registrations
from app.models import RegistrationTask
with get_db() as db:
failed_tasks = db.query(RegistrationTask).filter(
RegistrationTask.status == 'failed'
).all()
for task in failed_tasks:
print(f"Task {task.id}: {task.error_message}")
CPA Connection Issues
from app.services.cpa_service import CPAServiceManager
async def test_cpa_connection():
manager = CPAServiceManager()
try:
health = await manager.check_health(cpa_service_id=1)
print(f"CPA Status: {health['status']}")
except Exception as e:
print(f"Connection failed: {e}")
# Check base_url and admin_token in database
Account Pool Health Check
from app.services.inspection import InspectionService
async def health_check():
service = InspectionService()
result = await service.inspect_cpa_accounts(
cpa_service_id=1,
check_401=True,
check_quota=True,
auto_remove=False # Don't remove, just check
)
total = result['total_checked']
valid = result['valid_count']
health_percentage = (valid / total * 100) if total > 0 else 0
print(f"Pool health: {health_percentage:.1f}%")
if health_percentage < 50:
print("⚠️ Pool health critical, consider manual intervention")
Database Migration
# If using PostgreSQL, migrate from SQLite
from app.database import init_db
from app.models import Base
# Initialize new database schema
init_db()
# Manual data migration (example)
import sqlite3
from app.database import get_db
sqlite_conn = sqlite3.connect('data/database.db')
sqlite_cursor = sqlite_conn.cursor()
accounts = sqlite_cursor.execute(
"SELECT email, password, refresh_token, status FROM account_pool"
).fetchall()
with get_db() as db:
for email, password, token, status in accounts:
account = AccountPool(
email=email,
password=password,
refresh_token=token,
status=status
)
db.add(account)
db.commit()
Desktop Packaging
macOS
chmod +x scripts/build_macos_dmg.sh
./scripts/build_macos_dmg.sh
# Output: dist/CPA-Codex-Manager.dmg
Windows
scripts\build_windows.bat
REM Output: dist\CPA-Codex-Manager\CPA-Codex-Manager.exe
Environment Variables Reference
| Variable | Default | Description |
|---|---|---|
WEBUI_HOST | 127.0.0.1 | Web UI bind address |
WEBUI_PORT | 8000 | Web UI port |
WEBUI_ACCESS_PASSWORD | - | Admin password (required) |
APP_DATABASE_URL | data/database.db | Database connection string |
All secrets should be stored in environment variables or .env file, never hardcoded.
How can the creator link this skill?
Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.
<a href="https://skillzs.dev/skills/reason-machines/codex-skills/cpa-codex-manager-openai-pool">View cpa-codex-manager-openai-pool on skillZs</a>