ToolAtlas：通过工具端内存实现一次学习、处处复用
ToolAtlas: Learning Once, Reusing Everywhere with Tool-Side Memory
浏览论文内容
中文总结 AI 辅助
研究针对大语言模型代理依赖外部工具时工具知识难共享问题，提出ToolAtlas框架，通过执行验证探测构建提供商端工具内存，经实验证明该框架能有效提升性能且可跨环境和框架复用，确立了提供商端工具内存范式。
中文摘要 AI 辅助
大语言模型（LLM）代理越来越依赖共享提供商提供并由异构下游代理访问的外部工具。现有方法通过参数更新、提示细化或代理端内存来改进代理端的工具使用，使得工具知识难以共享且局限于过去任务中观察到的行为。我们认为可复用的工具知识应由工具提供商维护。我们引入了ToolAtlas，这是一个基于图的框架，通过执行验证的探测构建持久的提供商端工具内存，包括工具能力、失败边界和跨工具组合。在推理时，代理通过自适应图遍历查询工具内存。在跨越八项服务的两个基于MCP的基准测试中，ToolAtlas在pass@1上比现有工具端优化和代理端内存基线高出21.61%，在pass@4上高出18.61%。相同的工具内存还可跨环境实例和代理框架转移，无需重新训练或任务时探索，在pass@1/pass@4上分别产生高达24.16%/16.22%和17.49%/14.27%的相对增益。消融研究表明，这些增益源于将以工具为中心的内存组织与能力引导的执行探测相结合。这些结果确立了提供商端工具内存作为工具服务器的有效且可复用范式。
英文摘要
Large language model (LLM) agents increasingly rely on external tools served by shared providers and accessed by heterogeneous downstream agents. Existing approaches improve tool use on the agent side through parameter updates, prompt refinement, or agent-side memory, making tool knowledge difficult to share and limited to behaviors observed in past tasks. We argue that reusable tool knowledge should instead be maintained by the tool provider. We introduce ToolAtlas, a graph-based framework that builds a persistent provider-side tool memory of tool capabilities, failure boundaries, and cross-tool compositions through execution-verified probing. At inference time, agents query the tool memory via adaptive graph traversal. Across two MCP-based benchmarks spanning eight services, ToolAtlas outperforms existing tool-side optimization and agent-side memory baselines by up to 21.61% in pass@1 and 18.61% in pass@4. The same tool memory also transfers across environment instances and agent frameworks without retraining or task-time exploration, yielding up to 24.16%/16.22% and 17.49%/14.27% relative gains in pass@1/pass@4, respectively. Ablation studies show that these gains arise from combining tool-centered memory organization with capability-guided execution probing. These results establish provider-side tool memory as an effective and reusable paradigm for tool servers. Our code is in: https://github.com/PuppyKnightUniversity/ToolAtlas.
发表机构
- School of Computer Science, Peking University（北京大学计算机科学学院）
- Microsoft（微软）
机构由 AI 辅助整理，请以论文原文为准。
