mlsanityData to 7 Oct 2026

TaskCraft: Automated Generation of Agentic Tasks

arXiv 2506.10055 ↗First posted 11 Jun 2025cs.CLcited by 46

On the map: LLM agents and tool use → General LLM agent frameworks and benchmarks

Abstract (arXiv)

Agentic tasks, which require multi-step problem solving with autonomy, tool use, and adaptive reasoning, are becoming increasingly central to the advancement of NLP and AI. However, existing instruction data lacks tool interaction, and current agentic benchmarks rely on costly human annotation, limiting their scalability. We introduce textscTaskCraft, an automated workflow for generating difficulty-scalable, multi-tool, and verifiable agentic tasks with execution trajectories. TaskCraft expands atomic tasks using depth-based and width-based extensions to create structurally and hierarchically complex challenges. Empirical results show that these tasks improve prompt optimization in the generation workflow and enhance supervised fine-tuning of agentic foundation models. We present a large-scale synthetic dataset of approximately 36,000 tasks with varying difficulty to support future research on agent tuning and evaluation.

Key facts. Cited by 46 papers in the map (data through 7 Oct 2026). Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use. Builds on 26 of its 39 references that are in the map. First posted on arXiv 11 Jun 2025.

Builds on

26 of 39 references.

Closest ancestors · 5

Same direction, newest first

  1. 2025Agent models: Internalizing Chain-of-Action Generation into Reasoning models9 Mar
  2. 2024Magentic-One: A Generalist Multi-Agent System for Solving Complex Tasks7 Nov
  3. 2024Symbolic Learning Enables Self-Evolving Agents26 Jun
  4. 2023Agents: An Open-source Framework for Autonomous Language Agents14 Sep
  5. 2023CAMEL: Communicative Agents for "Mind" Exploration of Large Language Model Society31 Mar

Other references · 21

  1. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  2. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,686
  3. 2022Self-Instruct: Aligning Language Models with Self-Generated Instructionscited by 2,674
  4. 2023AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversationcited by 1,906
  5. 2025Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learningcited by 1,371
  6. 2023GAIA: a benchmark for General AI Assistantscited by 1,084
  7. 2022Measuring and Narrowing the Compositionality Gap in Language Modelscited by 1,054
  8. 2022Interleaving Retrieval with Chain-of-Thought Reasoning for Knowledge-Intensive Multi-Step Questionscited by 911
  9. 2021MuSiQue: Multihop Questions via Single-hop Question Compositioncited by 893
  10. 2023DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelinescited by 803
  11. 2023MetaMath: Bootstrap Your Own Mathematical Questions for Large Language Modelscited by 724
  12. 2025BrowseComp: A Simple Yet Challenging Benchmark for Browsing Agentscited by 616
9 more
  1. 2024Executable Code Actions Elicit Better LLM Agentscited by 537
  2. 2022LLM-Planner: Few-Shot Grounded Planning for Embodied Agents with Large Language Modelscited by 459
  3. 2023Tool Learning with Foundation Modelscited by 302
  4. 2024OpenMathInstruct-2: Accelerating AI for Math with Massive Open-Source Instruction Datacited by 194
  5. 2024MAmmoTH2: Scaling Instructions from the Webcited by 143
  6. 2023RecurrentGPT: Interactive Generation of (Arbitrarily) Long Textcited by 84
  7. 2024AutoAct: Automatic Agent Learning from Scratch for QA via Self-Planningcited by 81
  8. 2025AutoAgent: A Fully-Automated and Zero-Code Framework for LLM Agentscited by 49
  9. 2020MADER: Trajectory Planner in Multi-Agent and Dynamic Environmentscited by 21

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 18; 2026: 28.

46 papers.

Sort:

By area:

  1. 2025A Survey of Context Engineering for Large Language Modelscited by 120
  2. 2025WebShaper: Agentically Data Synthesizing via Information-Seeking Formalizationcited by 118
  3. 2025MemEvolve: Meta-Evolution of Agent Memory Systemscited by 81
  4. 2025Chain-of-Agents: End-to-End Agent Foundation Models via Multi-Agent Distillation and Agentic RLcited by 75
  5. 2025MiroThinker: Pushing the Performance Boundaries of Open-Source Research Agents via Model, Context, and Interactive Scalingcited by 72
  6. 2026SkillX: Automatically Constructing Skill Knowledge Bases for Agentscited by 69
  7. 2026AOrchestra: Automating Sub-Agent Creation for Agentic Orchestrationcited by 30
  8. 2025Flash-Searcher: Fast and Effective Web Agents via DAG-Based Parallel Executioncited by 27
  9. 2025AutoEnv: Automated Environments for Measuring Cross-Environment Agent Learningcited by 18
  10. 2026Self-Improvements in Modern Agentic Systems: A Surveycited by 17
  11. 2025Towards Personalized Deep Research: Benchmarks and Evaluationscited by 14
  12. 2026HunyuanOCR-1.5: Making Lightweight OCR VLMs Faster and Bettercited by 14
34 more
  1. 2025Environment Scaling for Interactive Agentic Experience Collection: A Surveycited by 13
  2. 2026EvoRoute: Experience-Driven Self-Routing LLM Agent Systemscited by 13
  3. 2026Terminal-World: Scaling Terminal-Agent Environments via Agent Skillscited by 13
  4. 2025AgentFrontier: Expanding the Capability Frontier of LLM Agents with ZPD-Guided Data Synthesiscited by 12
  5. 2025MLE-Smith: Scaling MLE Tasks with Automated Multi-Agent Pipelinecited by 9
  6. 2025WebAggregator: Enhancing Compositional Reasoning Capabilities of Deep Research Agent Foundation Modelscited by 9
  7. 2025CuES: A Curiosity-driven and Environment-grounded Synthesis Framework for Agentic RLcited by 8
  8. 2026Agentic Environment Engineering for Large Language Models: A Survey of Environment Modeling, Synthesis, Evaluation, and Applicationcited by 8
  9. 2025CRMWeaver: Building Powerful Business Agent via Agentic RL and Shared Memoriescited by 5
  10. 2026Search More, Think Less: Rethinking Long-Horizon Agentic Search for Efficiency and Generalizationcited by 4
  11. 2026The World Won't Stay Still: Programmable Evolution for Agent Benchmarkscited by 4
  12. 2026NexForge: Scaling Agent Capabilities through Requirement-Driven Task Synthesis for LLMscited by 4
  13. 2025Synthesizing Agentic Data for Web Agents with Progressive Difficulty Enhancement Mechanismscited by 3
  14. 2026ToolVerse: Unlocking Massive Environments and Long-Horizon Tasks for Agentic Reinforcement Learningcited by 3
  15. 2026MagicAgent: Towards Generalized Agent Planningcited by 2
  16. 2026Code2Math: Can Your Code Agent Effectively Evolve Math Problems Through Exploration?cited by 2
  17. 2026OmnilingualGAIA2: Evaluating the Multilingual Gap in Frontier AI Agentscited by 2
  18. 2026Apodex 1.1: Scaling Agentic Intelligence for Complex Workcited by 2
  19. 2025UniCode: Augmenting Evaluation for Code Reasoningcited by 1
  20. 2025Syn-GRPO: Self-Evolving Data Synthesis for MLLM Perception Reasoningcited by 1
  21. 2026Beyond Accuracy: A Cognitive Load Framework for Mapping the Capability Boundaries of Tool-use Agentscited by 1
  22. 2026TodoEvolve: Learning to Architect Agent Planning Systemscited by 1
  23. 2026QuarkMedSearch: A Long-Horizon Deep Search Agent for Exploring Medical Intelligencecited by 1
  24. 2026Scalable Environments Drive Generalizable Agentscited by 1
  25. 2026Efficient Agentic Reasoning Through Self-Regulated Simulative Planningcited by 1
  26. 2026ISE: An Execution-Grounded Recipe for Multi-Turn OS-Agent Trajectoriescited by 1
  27. 2026State2State: Environment-Derived Mid-Training for LLM Agentscited by 1
  28. 2025Anemoi: A Semi-Centralized Multi-agent System Based on Agent-to-Agent Communication MCP server from Coral Protocolcited by 0
  29. 2026WebClipper: Efficient Evolution of Web Agents with Graph-based Trajectory Pruningcited by 0
  30. 2026Project Auto-World: Towards Automated Benchmarking of Neural Relational Reasonerscited by 0
  31. 2026UnifiedPlayers: Enhance Tool-Integrated Reasoning in Agentic Reinforcement Learningcited by 0
  32. 2026AutoDataBench: Can Agents Write the Data That Feeds the Self-Improvement Loop?cited by 0
  33. 2026WEFT: Scaling Tool-Use Post-Training for General-Purpose Agentscited by 0
  34. 2026SkillGym: Training Skill-Use Agents with Automatic Verifiable Environment Generationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025AgentSynth: Scalable Task Generation for Generalist Computer-Use Agents
  2. 2026A Comprehensive Survey on Agent Skills: Taxonomy, Techniques, and Applications
  3. 2026AgentFactory: Towards Automated Agentic System Design and Optimization
  4. 2024Advancing Agentic Systems: Dynamic Task Decomposition, Tool Integration and Evaluation using Novel Metrics and Dataset
  5. 2026DIVE: Scaling Diversity in Agentic Task Synthesis for Generalizable Tool Use
  6. 2024AgentQuest: A Modular Benchmark Framework to Measure Progress and Improve LLM Agents
  7. 2026Meta-Task: Turning Terminal Task Synthesis into a Terminal Task for Scalable Agent Training
  8. 2025FABRIC: Framework for Agent-Based Realistic Intelligence Creation