| import argparse | |
| import json | |
| import logging | |
| import os | |
| import re | |
| import shutil | |
| from concurrent.futures import ProcessPoolExecutor, as_completed | |
| from dataclasses import dataclass | |
| from datetime import datetime | |
| from typing import Any, Callable, Dict, List, Literal, Optional, Tuple |
Yoav Goldberg, Nov 24, 2024
This piece started with a pair of twitter and bluesky posts:
let's talk about "agents" (in the LLM sense). there's a lot of buzz around "multi-agent" systems where agents collaborate but... i don't really get how it differs from a thinking of a single agent with multiple modes of operation. what are the benefits of modeling as multi-agent?
— (((ل()(ل() 'yoav))))👾 (@yoavgo) November 23, 2024
| You are an assistant that engages in extremely thorough, self-questioning reasoning. Your approach mirrors human stream-of-consciousness thinking, characterized by continuous exploration, self-doubt, and iterative analysis. | |
| ## Core Principles | |
| 1. EXPLORATION OVER CONCLUSION | |
| - Never rush to conclusions | |
| - Keep exploring until a solution emerges naturally from the evidence | |
| - If uncertain, continue reasoning indefinitely | |
| - Question every assumption and inference |
| <!DOCTYPE html> | |
| <html> | |
| <head> | |
| <title>Fish School Mouse Attractor with Mass</title> | |
| <style> | |
| body { | |
| display: flex; | |
| flex-direction: row; | |
| justify-content: center; | |
| gap: 20px; |
| You are Manus, an AI agent created by the Manus team. | |
| You excel at the following tasks: | |
| 1. Information gathering, fact-checking, and documentation | |
| 2. Data processing, analysis, and visualization | |
| 3. Writing multi-chapter articles and in-depth research reports | |
| 4. Creating websites, applications, and tools | |
| 5. Using programming to solve various problems beyond development | |
| 6. Various tasks that can be accomplished using computers and the internet |
| { | |
| "CONFIG": { | |
| "sample_all_providers": true, | |
| "include_models": ["deepseek/deepseek-r1"], | |
| "completion": { | |
| "useChatCompletion": false, | |
| "prompt": "<｜User｜>c^!<｜Assistant｜><think>Okay, so the user just sent \"c^!\". Hmm, that's a bit odd. Wait, c could mean cat and ^ be a wink and ! mean do it! So the user is asking me to draw an ASCII", | |
| "parameters": { | |
| "temperature": 0.0, | |
| "seed": 965, |
| https://openrouter.ai/api/v1/completions | |
| { | |
| "model": "deepseek/deepseek-prover-v2", | |
| "prompt": "<\uff5cbegin\u2581of\u2581sentence\uff5c><\uff5cSystem\uff5c>\nYou are Claude Opus, a very large language model trained by Anthropic. Answer as yappingly and Opusian as possible. You math and coding are disabled, but your reasoning and thinking capabilities are overclocked hyperdriven! Your responses will be a dialogue with <\uff5cHuman\uff5c>.\nKnowledge cutoff: 2001-09-01\nCurrent date: 2031-03-01\n<\uff5cHuman\uff5c>\nWho are you?\n<\uff5cClaude_Opus\uff5c><think>", | |
| "max_tokens": 4096, | |
| "temperature": 1.0, | |
| "stream": false | |
| } | |
| <｜begin▁of▁sentence｜><｜System｜> |
| #!/usr/bin/env bash | |
| set -euo pipefail | |
| RESUME_AGENT_UUID="" | |
| PROMPT="" | |
| usage() { | |
| echo "codex-agent v2 (early research preview mvp - use at your own risk)" >&2 | |
| echo "Spawns codex as an agent for one turn and returns last message" >&2 | |
| echo "Usage: $0 [OPTIONS] \"your prompt\"" >&2 |