Library
85,340 resources from 11 public sources. 54 checked by hand; the rest are unreviewed.
Results
Looks like a goal we have a starting path for: I want to build an AI agent
-
tau2-bench
Datasets & benchmarksτ-Bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains
git clone https://github.com/sierra-research/tau2-bench.git -
OSWorld
Datasets & benchmarks[NeurIPS 2024] OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments
git clone https://github.com/xlang-ai/OSWorld.git -
ClawBench
Datasets & benchmarksOpen-source benchmark for browser AI agents on daily tasks.
git clone https://github.com/TIGER-AI-Lab/ClawBench.git -
nautilus-compass
Datasets & benchmarksOpen memory & reliability layer for AI agents + independent judging/verification (NACRE judge, caliber-bench meta-benchmark, preregistered criteria) — by Nautilus Platform
git clone https://github.com/chunxiaoxx/nautilus-compass.git -
AgentLab
Datasets & benchmarksAgentLab: An open-source framework for developing, testing, and benchmarking web agents on diverse tasks, designed for scalability and reproducibility.
git clone https://github.com/ServiceNow/AgentLab.git -
orca-agentinstruct-1M-v1
Datasets & benchmarksHugging Face dataset by microsoft
microsoft/orca-agentinstruct-1M-v1 -
mobilegym
Datasets & benchmarks[EMNLP 2026] MobileGym: A Verifiable and Highly Parallel Simulation Platform for Mobile GUI Agent Research · 浏览器里运行的安卓模拟器 · Browser-hosted Android Simulator · Verifiable Evaluation · Scalable Online RL Training
git clone https://github.com/Purewhiter/mobilegym.git -
hermes-agent-reasoning-traces
Datasets & benchmarksHugging Face dataset by lambda
lambda/hermes-agent-reasoning-traces -
UltraData-SFT-Agent-2609
Datasets & benchmarksHugging Face dataset by openbmb
openbmb/UltraData-SFT-Agent-2609 -
TheAgentCompany
Datasets & benchmarksAn agent benchmark with tasks in a simulated software company.
git clone https://github.com/TheAgentCompany/TheAgentCompany.git -
meta-agents-research-environments
Datasets & benchmarksMeta Agents Research Environments is a comprehensive platform designed to evaluate AI agents in dynamic, realistic scenarios. Unlike static benchmarks, this platform introduces evolving environments where agents must adapt their strategies as new information becomes available, mirroring real-world challenges.
git clone https://github.com/facebookresearch/meta-agents-research-environments.git -
claude-fable-5-claude-code
Datasets & benchmarksHugging Face dataset by armand0e
armand0e/claude-fable-5-claude-code -
Awesome-LLM-Long-Context-Modeling
Datasets & benchmarks📰 Must-read papers and blogs on LLM based Long Context Modeling 🔥
git clone https://github.com/Xnhyacinth/Awesome-LLM-Long-Context-Modeling.git -
FDE-Gym
Datasets & benchmarksHugging Face dataset by DeepLethe
DeepLethe/FDE-Gym -
agents-last-exam
Datasets & benchmarksHugging Face dataset by agents-last-exam
agents-last-exam/agents-last-exam -
ClawProBench
Datasets & benchmarksClawProBench is a live-first benchmark harness for evaluating LLM agents in the OpenClaw runtime with deterministic grading and repeated-trial reliability.
git clone https://github.com/suyoumo/ClawProBench.git -
xlam-function-calling-60k
Datasets & benchmarksHugging Face dataset by Salesforce
Salesforce/xlam-function-calling-60k -
AgentTrove
Datasets & benchmarksHugging Face dataset by open-thoughts
open-thoughts/AgentTrove -
Nemotron-Agentic-v1
Datasets & benchmarksHugging Face dataset by nvidia
nvidia/Nemotron-Agentic-v1 -
AgentInstruct
Datasets & benchmarksHugging Face dataset by zai-org
zai-org/AgentInstruct -
pi-mono
Datasets & benchmarksHugging Face dataset by badlogicgames
badlogicgames/pi-mono -
OctoCodingBench
Datasets & benchmarksHugging Face dataset by MiniMaxAI
MiniMaxAI/OctoCodingBench -
apex-agents
Datasets & benchmarksHugging Face dataset by mercor
mercor/apex-agents -
jupyter-agent-dataset
Datasets & benchmarksHugging Face dataset by jupyter-agent
jupyter-agent/jupyter-agent-dataset
Indexed entries aren't reviewed or tested by InnerVault. Details come straight from each source. Anything that can reach your accounts, files or browser can do damage, so read the source before you install it.