Formal Skill: Programmable Runtime Skills for Efficient and Accurate LLM Agents

arXiv cs.AI Papers
llm-agents runtime-skills formal-skills programmable-runtime event-driven open-source arxiv

Summary

This paper introduces Formal Skill, a runtime-native abstraction for LLM agents that encodes reusable procedures as executable state machines with JSON metadata, Python executors, and hook-governed control logic. An open-source implementation called FairyClaw is presented, showing competitive performance on Harness-Bench with reduced token usage.

arXiv:2605.19604v1 Announce Type: new Abstract: Large Language Model (LLM) agents increasingly act inside real workspaces, where tools and skills determine whether model reasoning becomes reliable action. Existing skills remain largely informal: Markdown skills and instruction packs encode procedures as long natural-language documents, while function calling, Model Context Protocol (MCP) servers, and framework tools structure individual actions but usually leave workflow state, policy enforcement, and completion discipline outside the skill itself. We introduce Formal Skill, a runtime-native abstraction that represents reusable capability with JSON metadata and action schemas, reliable Python executors, hook-governed control logic, Formal Skill routing, and skill-local runtime state. By moving reusable procedure from repeated prompt text into executable state machines and hook policies, Formal Skill gives agents a token-efficient and enforceable control surface. We implement the abstraction in FairyClaw, an open-source event-driven runtime for executable, observable, and composable Formal Skills. On Harness-Bench, FairyClaw obtains highly competitive average scores while using substantially fewer tokens, with especially strong results on tasks that expose the role of Formal Skill.
Original Article

Similar Articles

SKILL.state: Scalable Long-Horizon Agent Skills

arXiv cs.AI

SKILL.state introduces a runtime architecture for LLM-based agents that uses mutable execution state instead of growing conversation history, enhancing accuracy and reducing token usage in long-horizon tasks.

Harnessing LLM Agents with Skill Programs

Hugging Face Daily Papers

HASP is a framework that upgrades agent skills into executable program functions acting as guardrails, enabling direct intervention in LLM agent loops and improving performance on complex tasks like web-search, math reasoning, and coding.

Better, Faster, Stronger: Programmatic Skill Learning Best Reduces Agent Cost

arXiv cs.CL

This paper argues that representing agent skills as code, rather than natural language, yields the best cost reduction for LLM agents, and introduces SpeedRunner, a coding agent that learns programmatic skills from past trajectories to improve performance while cutting costs across embodied environments.