MIND-Skill: Quality-Guaranteed Skill Generation via Multi-Agent Induction and Deduction

2026-05-09

AI Summary

This work addresses the limitation of large language model (LLM) agents in executing complex, multi-step tasks requiring domain-specific procedural knowledge, primarily due to a lack of reusable skills and heavy reliance on human experts. To overcome this, the authors propose the MIND-Skill framework, which employs an inductive agent to automatically abstract skills from successful execution trajectories and a deductive agent to validate their efficacy through trajectory reconstruction. The approach introduces a novel tripartite constraint—comprising reconstruction loss, outcome loss, and rubric-based loss—jointly optimized with TextGrad to ensure skill correctness, generalization, and high-quality documentation. Evaluated on the AppWorld and BFCL-v3 benchmarks, MIND-Skill significantly outperforms existing methods, demonstrating superior generalization and task-completion performance on unseen tasks.

Abstract

Large language model (LLM) powered AI agents have emerged as a promising paradigm for autonomous problem-solving, yet they continue to struggle with complex, multi-step real-world tasks that demand domain-specific procedural knowledge. Reusable agent skills, which encapsulate successful problem-solving strategies, offer a natural remedy by enabling agents to build on prior experience. However, curating such skills has largely remained a manual endeavor, requiring human experts to distill rich domain knowledge into actionable guidelines. In this work, we present $\textbf{M}$ulti-agent $\textbf{IN}$duction and $\textbf{D}$eduction for $\textbf{Skill}$s ($\textbf{MIND-Skill}$), a framework that automatically induces generalizable skills from successful trajectories with robust quality guarantees. MIND-Skill consists of an induction agent which is tasked to abstract reusable skills from successful trajectories, and a deduction agent which aims to reconstruct trajectories by following the induced skills. To guarantee the quality of the generated skills, we introduce a reconstruction loss that compares input and reconstructed trajectories, an outcome loss that enforces the correctness of the reconstructed trajectories, and a rubric loss that assesses the documentation quality and regularizes the abstraction level of the generated skills according to predefined criteria. These textual losses are jointly optimized with TextGrad, and the resulting skills are evaluated on held-out tasks unseen during optimization. Experiments on AppWorld and BFCL-v3 show that MIND-Skill consistently outperforms concurrent skill generation methods.

Problem

• skill generation
• autonomous problem-solving
• procedural knowledge
• quality guarantee
• multi-agent framework

Innovation

• multi-agent induction and deduction
• skill generation
• reconstruction loss
• TextGrad
• quality-guaranteed abstraction