pa-status（PA 状态）

PA Status Skill

Minimum Model

Any model. Status checks are data-driven and rule-based.

When to Run

• Daily at 09:00: Full network report to admin.
• Immediately on billing error: Send a partial report for the affected PA.
• On demand: If admin asks "what's the status?" → generate the report now.

Status Checks

For each PA in data/pa-directory.json, check:

Automated Report Script

#!/usr/bin/env python3
import json
import datetime

# Load PA directory
try:
    with open('data/pa-directory.json') as f:
        d = json.load(f)
except FileNotFoundError:
    print("ERROR: data/pa-directory.json not found")
    exit(1)

today = datetime.date.today().isoformat()
online = []
issues = []
offline = []

for pa in d.get('pas', []):
    name = pa['name']
    owner = pa['owner']
    status = pa.get('status', 'unknown')
    model = pa.get('model', 'unknown')

    # Format calendar and billing indicators
    calendar = "✅" if pa.get('calendar_connected') else "❌"
    billing_ok = not pa.get('billing_error', False)
    billing = "✅" if billing_ok else "⚠️ billing error"

    # Classify each PA into online / issues / offline
    if status == 'active' and billing_ok:
        online.append(f"• {name} ({owner}) — {model}, cal {calendar}")
    elif status == 'inactive':
        offline.append(f"• {name} ({owner})")
    else:
        issues.append(f"• {name} ({owner}) — {billing}")

# Print the report
total = len(d.get('pas', []))
print(f"📊 PA Network Status — {today}\
")
print(f"✅ ONLINE ({len(online)}/{total})")
for line in online:
    print(line)

if issues:
    print("\
⚠️ ISSUES")
    for line in issues:
        print(line)

if offline:
    print("\
❌ OFFLINE")
    for line in offline:
        print(line)

if not issues and not offline:
    print("\
All PAs are healthy 🎉")

Directory Schema (With Status Fields)

Add these fields to each PA entry in pa-directory.json:

{
  "name": "Aria",
  "phone": "+1XXXXXXXXXX",
  "owner": "Owner Name",
  "owner_email": "owner@company.com",
  "status": "active",
  "model": "your-llm-model",
  "last_seen": "2026-04-01T10:00:00Z",
  "calendar_connected": true,
  "billing_error": false,
  "billing_error_since": null
}

Replace "your-llm-model" with the actual model ID (e.g. "claude-haiku-20240307", "gpt-4o-mini", "gemini-1.5-flash").

Quick Ping (WhatsApp Reachability)

Use this for real-time check (after the script above):

For each PA in the issues list:
  Send: "ping 🏓"
  Wait up to 5 minutes
  If response received: mark ONLINE
  If no response: mark NO_RESPONSE → check whatsapp-diagnostics and billing-monitor skills

Rule: Only ping PAs that are flagged. Do not ping healthy PAs — it creates noise.

Scheduling

Cost Tips

• Very cheap: Reading a JSON file and formatting a report uses minimal tokens.
• Small model OK: Any model can generate this report.
• Avoid: Do not ping all PAs at once — send one at a time to avoid rate limits.
• Batch: Run the full report once daily instead of checking each PA separately.