Skip to content

Instantly share code, notes, and snippets.

View dudutwizer's full-sized avatar

David (Dudu) Twizer dudutwizer

View GitHub Profile
@dudutwizer
dudutwizer / response.json
Created March 5, 2026 03:16
response.json
[
{
"url": "https://api.github.com/repos/openclaw/openclaw/issues/35306",
"repository_url": "https://api.github.com/repos/openclaw/openclaw",
"labels_url": "https://api.github.com/repos/openclaw/openclaw/issues/35306/labels{/name}",
"comments_url": "https://api.github.com/repos/openclaw/openclaw/issues/35306/comments",
"events_url": "https://api.github.com/repos/openclaw/openclaw/issues/35306/events",
"html_url": "https://github.com/openclaw/openclaw/issues/35306",
"id": 4025419689,
"node_id": "I_kwDOQb6kR87v7wep",
@dudutwizer
dudutwizer / self_improving_agent.py
Last active January 11, 2026 03:34
Self-Improving Agent: Learns from execution history (past tasks/threads) and uses Xpander API to update its own system prompt based on what it learned
from typing import Any, Dict, List, Optional, Tuple
import os
from dotenv import load_dotenv
import requests
from xpander_sdk import Backend, Task, Tokens, on_task, register_tool
from agno.agent import Agent
load_dotenv()
@dudutwizer
dudutwizer / app.py
Created January 10, 2026 01:55
Python script for parallel webhook calls with timing proof
import asyncio
import time
from datetime import datetime
import aiohttp
async def call_webhook(session, call_id):
"""Make a single webhook call and return timing info."""
url = 'https://webhook.xpander.ai?agent_id=YOUR_AGENT_ID'
headers = {
@dudutwizer
dudutwizer / README.md
Last active September 1, 2025 22:21
Agno AI Agent Workflow - Task Management System

Agno AI Agent Workflow - Forced Task Completion with Proof

The Problem

AI agents often give vague responses and stop before fully completing tasks.

The Solution

This workflow forces agents to:

  1. Break requests into specific subtasks
  2. Complete each task with concrete evidence
  3. Get human approval before marking anything "done"
from xpander_sdk import XpanderClient, LLMProvider # pip install https://assets.xpanderai.io/xpander-sdk.tar.gz
from openai import OpenAI # pip install openai
from langchain_nvidia_ai_endpoints import ChatNVIDIA # pip install langchain_nvidia_ai_endpoints
# Loads all the models that supports Function calling
tool_models = [model for model in ChatNVIDIA.get_available_models() if model.supports_tools]
xpander_client = XpanderClient(
agent_key="<changeme>", #Get it from https://app.xpander.ai
agent_url="https://inbound.xpander.ai/agent/<change-me>", #Get it from https://app.xpander.ai