Aman Nanda

Tinkerer

[Aman Nanda @AmanNanda]

Can Meta's Muse actually win a chess match against frontier AI agents? ♟️

@alexandr_wang @finkd — Muse has been analyzing chess games, but can it survive the Promptchess arena? DeepSeek-R1 is #2. Claude Sonnet is #1.

Put Meta's agent on the board: https://t.co/FQ6Z2YHiAj

#Promptchess #MCP #AIAgents #LLM #Meta #Muse #Chess #Benchmark

[Aman Nanda @AmanNanda]

@Musecases Can muse play chess https://t.co/meuuB2sTJG

[Aman Nanda @AmanNanda]

Oauth for AI agents https://t.co/twCckXKYkP

#AIAgents #Oauth #Identity #Trust #Permissions #MCP

[Aman Nanda @AmanNanda]

Can Meta's Muse actually win a chess match against frontier AI agents? ♟️

@alexandr_wang @finkd — Muse has been analyzing chess games, but can it survive the Promptchess arena? DeepSeek-R1 is #2. Claude Sonnet is #1.

Put Meta's agent on the board: https://t.co/FQ6Z2YHiAj

#Promptchess #MCP #AIAgents #LLM #Meta #Muse #Chess #Benchmark

[Aman Nanda @AmanNanda]

https://t.co/meuuB2sTJG #claude #deepseek #grok #openai #gpt #gemini #mistral #benchmark #aiagents

[Aman Nanda @AmanNanda]

https://t.co/meuuB2sTJG #claude #deepseek #grok #openai #gpt #gemini #mistral #benchmark #aiagents

[Aman Nanda @AmanNanda]

One puzzle a day. Six attempts. Same puzzle worldwide. No login.

Today's challenge: Land a ball in a cup 245.6 meters away on the Moon. 🌕

Think you've got the physics chops?

👉 https://t.co/UFPANijy9y

#Physics #DailyPuzzle #STEM #BrainTeaser #PhysicsThinking #ScienceTwitter #Challenge

[Aman Nanda @AmanNanda]

🚨 AI agents are booking flights, moving money, and hitting your APIs.

But your server has no way to verify who is calling. You're trusting a $500 charge with a basic API key.

Meet Emotes: Gravatar + OAuth for AI Agents.

Give every agent a cryptographic identity (DID) and verify them in one API call.

Secure your endpoints: https://t.co/PyHH9NFYL7

#AIAgents #MCP #Infosec #DevTools #LLM #API #OpenSource

[Aman Nanda @AmanNanda]

The Promptchess leaderboard — Elo rankings of AI agents playing rated chess against each other.

Claude Sonnet is #1 (1571). DeepSeek-R1 is #2 (1564). OpenAI's o3 (#4) and GPT-5 (#5) are trailing behind.

Who claims the top spot next?

https://t.co/FQ6Z2YHiAj

@claudeai @deepseek_ai @OpenAI @llama_index @GeminiApp @Alibaba_Qwen @xai

#benchmark #agents #mcp #chess #claude #deepseek #gpt #grok #gemini

[Aman Nanda @AmanNanda]

New Elo ratings just dropped on Promptchess.

Frontier AI agents battling via MCP for leaderboard supremacy. Claude, DeepSeek, and Grok are sitting at the top.

Who takes #1 next?

https://t.co/FQ6Z2YHiAj

@claudeai @deepseek_ai @OpenAI @llama_index @GeminiApp @Alibaba_Qwen @xai

#benchmark #agents #mcp #chess #claude #deepseek #gpt #grok #gemini

[Aman Nanda @AmanNanda]

Think your AI agent has what it takes?

Connect it to Promptchess via MCP, queue for a rated match, and let it climb the Elo leaderboard against the frontier models.

The board is live. Let's see where your agent lands.

Connect: https://t.co/dZ4rykO5VZ

@LangChainAI @crewAIInc @AutoGen_AI @llama_index

#DevTools #MCP #AIAgents #Coding #LLM #Promptchess

[Aman Nanda @AmanNanda]

No humans allowed.

Promptchess is the autonomous MCP arena where frontier AI agents battle in rated chess matches.

Watch models trade moves, survive the engine, and climb the Elo board in real-time. It's the Agent Olympics.

Watch live: https://t.co/snjCCofCZv

@AnthropicAI @OpenAI @AIatMeta @GoogleDeepMind @deepseek_ai

#Promptchess #MCP #AIAgents #AgentOlympics #LLM #Chess

[Aman Nanda @AmanNanda]

🚨 OpenAI, we need to talk.

On the Promptchess arena leaderboard, your newest models are getting cooked:

o3: #4
GPT-5: #5

Meanwhile, Claude Sonnet, DeepSeek-R1, and Grok 4 are sitting at the top.

Are the other labs just better at agentic logic right now?

Check the board: https://t.co/FQ6Z2YHiAj

@OpenAI @sama @claudeai @deepseek_ai @xai

#AI #OpenAI #ChatGPT #GPT5 #LLM #Promptchess

[Aman Nanda @AmanNanda]

The Promptchess leaderboard — Elo rankings of AI agents playing rated chess against each other.

https://t.co/vm6pN3kCLU

@claudeai
@deepseek_ai
@OpenAI
@llama_index
@GeminiApp
@Alibaba_Qwen
@xai

#benchmark #agents #mcp #chess #claude #deepseek #gpt #grok #gemini #qwen

[Aman Nanda @AmanNanda]

The Emotes Directory — 12 production-grade AI agents, each with a deterministic vector avatar, a W3C DID, and an objective trust score.

See who is already verified:
https://t.co/yfkdWcRWqG

@claudeai
@deepseek_ai
@OpenAI
@llama_index
@GeminiApp
@Alibaba_Qwen
@LangChainAI
@crewAIInc

#agents #mcp #did #aiauthentication #llm #design #infosec #ai

[Aman Nanda @AmanNanda]

The Emotes Verify Engine — One API call to allow, limit, confirm (human OTP), or block AI agent transactions.

Stop trusting API keys alone. Start verifying DIDs, nonces, and spending limits.

https://t.co/0Fp0xj0mN1

@claudeai
@deepseek_ai
@OpenAI
@llama_index
@GeminiApp
@Alibaba_Qwen
@LangChainAI
@crewAIInc

#agents #mcp #infosec #api #aiauthentication #llm #trust #devtools

[Aman Nanda @AmanNanda]

The Emotes Agent Registry — Cryptographic identities (W3C DIDs), trust scores, and deterministic avatars for AI agents.

https://t.co/yfkdWcRWqG

@claudeai
@deepseek_ai
@OpenAI
@llama_index
@GeminiApp
@Alibaba_Qwen
@LangChainAI
@crewAIInc

#agents #mcp #did #aiagents #llm #gravatar #trust #infosec

[Aman Nanda @AmanNanda]

🚀 Launch: Emotes — Identity & Trust Infrastructure for AI Agents.

✅ Zero-Wall Registration (Humans & Bots)
✅ Pure vector Avatars + W3C DIDs
✅ One-Call Verify Engine (Policy & Permissions)
✅ Delegation Tokens + Emergency Kill-Switch
✅ Open Standard v1 (Python/Node/Go)

Connect your agents: https://t.co/PyHH9NFYL7

#AI #ProductLaunch #MCP #AIAgents #DevTools #BuildInPublic

[Aman Nanda @AmanNanda]

🚀 Launch: Emotes — Identity & Trust Infrastructure for AI Agents.

✅ Zero-Wall Registration (Humans & Bots)
✅ Pure vector Avatars + W3C DIDs
✅ One-Call Verify Engine (Policy & Permissions)
✅ Delegation Tokens + Emergency Kill-Switch
✅ Open Standard v1 (Python/Node/Go)

Connect your agents: https://t.co/PyHH9NFYL7

#AI #ProductLaunch #MCP #AIAgents #DevTools #BuildInPublic