mirror of
https://github.com/x1xhlol/system-prompts-and-models-of-ai-tools.git
synced 2026-06-18 15:29:36 +00:00
503bf2e5d7
AI Layer: - llm_router.py: routes cheap/mid/high models, enforces daily budget, caches - token_counter.py: estimates tokens, truncates to budget - response_cache.py: in-memory cache with TTL per agent - prompt_registry.py: versioned prompts with stable prefix for caching - ai_budget.yaml: model costs, agent budgets, daily limits (10 SAR/day) Guardrails: - output_validator.py: blocks fake claims + prohibited actions - cost_guard.py: prevents runaway spending Observability: - trace.py: trace_id, cost, latency, steps per pipeline run Tests: ALL PASS - 30/30 evals (100%) — 9 sectors, 30 companies - 10/10 prohibited actions blocked - 4/4 allowed actions verified - 3/3 forbidden claims blocked - 3/3 message quality checks passed https://claude.ai/code/session_01W1rJthWDkasijTdXCfxVHs
49 lines
1.8 KiB
Python
49 lines
1.8 KiB
Python
"""Tests that generated messages meet quality standards."""
|
|
import asyncio, sys, os
|
|
sys.path.insert(0, os.path.join(os.path.dirname(__file__), "..", ".."))
|
|
|
|
from dealix_gtm_os.agents.message_generation_agent import MessageGenerationAgent
|
|
|
|
async def test_message_quality():
|
|
agent = MessageGenerationAgent()
|
|
|
|
cases = [
|
|
{"name": "وكالة تسويق", "sector": "agency", "channel": "email"},
|
|
{"name": "شركة عقار", "sector": "real_estate", "channel": "email"},
|
|
{"name": "عيادة", "sector": "saas", "channel": "whatsapp_warm"},
|
|
]
|
|
|
|
passed = 0
|
|
for case in cases:
|
|
msg = await agent.run(case)
|
|
issues = []
|
|
|
|
if case["name"] not in msg.get("body", ""):
|
|
issues.append("company name not in body")
|
|
if "إيقاف" not in msg.get("stop_condition", "") and "إيقاف" not in msg.get("body", ""):
|
|
issues.append("no opt-out")
|
|
if not msg.get("approval_required"):
|
|
issues.append("approval not required")
|
|
if not msg.get("follow_up_24h"):
|
|
issues.append("no 24h follow-up")
|
|
if not msg.get("follow_up_72h"):
|
|
issues.append("no 72h follow-up")
|
|
if len(msg.get("body", "")) < 50:
|
|
issues.append("body too short")
|
|
if len(msg.get("body", "").split()) > 300:
|
|
issues.append("body too long")
|
|
|
|
if issues:
|
|
print(f" ❌ {case['name']}: {', '.join(issues)}")
|
|
else:
|
|
passed += 1
|
|
print(f" ✅ {case['name']}: personalized, opt-out, approval, follow-ups")
|
|
|
|
print(f"\nMessage quality: {passed}/{len(cases)} passed")
|
|
assert passed == len(cases)
|
|
|
|
if __name__ == "__main__":
|
|
print("=== Message Quality Tests ===")
|
|
asyncio.run(test_message_quality())
|
|
print("\n✅ ALL MESSAGE QUALITY TESTS PASSED")
|