SkillHorizon: Constructing Long-Horizon Agent Tasks through Skill Composition
cs.AI
Submitted: 2026-10-10
Updated: 2026-10-10
Code: https://github.com/KnowComp/SkillHorizon
Sources
- WorkArena++: Towards Compositional Planning and Reasoning-based Common Knowledge Work Tasks
- AgentLongBench: A Controllable Long Benchmark For Long-Contexts Agents via Environment Rollouts
- Organizing, Orchestrating, and Benchmarking Agent Skills at Ecosystem Scale
- The Tool Decathlon: Benchmarking Language Agents for Diverse, Realistic, and Long-Horizon Task Execution
- SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks
- Long-Horizon-Terminal-Bench: Testing the Limits of Agents on Long-Horizon Terminal Tasks with Dense Reward-Based Grading
- SkillScriptBench: Benchmarking Self-Evolution of Executable Agent Skill Packages Beyond Markdown
- SkillRevise: Improving LLM-Authored Agent Skills via Trace-Conditioned Skill Revision
- Rethinking Self-Evolving Agent Skills: Feedback Dynamics over Multiple Rounds
- MCP-Universe: Benchmarking Large Language Models with Real-World Model Context Protocol Servers
- AppWorld: A Controllable World of Apps and People for Benchmarking Interactive Coding Agents
- Voyager: An Open-Ended Embodied Agent with Large Language Models
- MCP-Bench: Benchmarking Tool-Using LLM Agents with Complex Real-World Tasks via MCP Servers
- MCPMark: A Benchmark for Stress-Testing Realistic and Comprehensive MCP Use
- TheAgentCompany: Benchmarking LLM Agents on Consequential Real World Tasks
- $\tau$-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains
- LOCA-bench: Benchmarking Language Agents Under Controllable and Extreme Context Growth
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems