Tool use and MCP
17 notes, newest first. Includes the vault tags tool-use, mcp, model-context-protocol.
ExplorationBench: Measuring AI Systems' Exploration in Verifiable Alien Worlds
ExplorationBench evaluates whether AI systems can acquire genuinely unfamiliar rules through exploration rather than recall. It uses two executable sandboxes, AlienCode and AlienLogic, with discovery targets, flawed…
HEXIS: Compiling Skills into Extended Finite State Machines
HEXIS compiles reusable agent skills into extended finite state machines that separate task knowledge from control flow. The compiler maps skill clauses and tool interfaces to states, local instructions, data bindings…
Jev-Mobile: Jev as an Executor for Mobile GUI Agents
Jev-Mobile separates low-frequency vision-language-model planning from high-frequency lightweight execution. The VLM supplies local goals, the accessibility tree defines a structured action space, and Jev repeatedly…
Harness-Zero: Harness Distillation via Agent-as-Harness
Harness-Zero uses a domain- or instance-optimized harness as training-time guidance, then transfers its behavior into model weights so the specialized harness can be removed at deployment. An agent-as-harness corrects…
MidTool: Mid-training Data Synthesis for Agentic Tool Use
Researchers have developed MidTool, a system for teaching large language models how to use tools effectively during mid-training. This is done by combining large-scale web and PDF data with synthetic supervision from…
MidTool: Mid-training Data Synthesis for Agentic Tool Use
In the paper 'MidTool: Mid-training Data Synthesis for Agentic Tool Use', researchers present a new open corpus construction pipeline called MidTool. This pipeline is designed to teach large language models how to use…
Retry, Switch, or Abstain? Learning Strategy-Aware Tool-Use Policies via Controlled Error Injection
The paper presents a framework called BENCH2ROBUST, which converts reliable tool-use benchmarks into controlled stochastic environments with failure scenarios. The authors test two interventions to improve the…
Retry, Switch, or Abstain? Learning Strategy-Aware Tool-Use Policies via Controlled Error Injection
Researchers present BENCH2ROBUST, a framework for converting reliable tool-use benchmarks into stochastic environments with controlled solvability. This allows study of agent's response to tool failures, which can be…
A Gateway Architecture for Enterprise MCP Authentication: Unifying Heterogeneous Auth, Identity Delegation, and the User / Non-User Persona Problem
This research presents a solution to the governance crisis caused by the rapid adoption of Model Context Protocol (MCP) in large organizations. The proposed solution is a centralized MCP gateway that unifies…
OpenForgeRL: Train Harness-native Agents in Any Environment
OpenForgeRL is an open-source framework designed to train AI agents using complex inference harnesses like Claude Code, Codex, and OpenClaw, in various environments. The framework allows end-to-end training of these…
OpenForgeRL: Train Harness-native Agents in Any Environment
The paper introduces OpenForgeRL, an open-source framework for training AI agents using harnesses like Claude Code, Codex, and OpenClaw in various environments. It allows end-to-end training of these complex harnesses…
From Legacy Documentation to OSCAL: An MCP-Based Agent Pipeline for Threat-Informed Continuous Compliance in Critical Infrastructure
The paper proposes an agent pipeline using Machine Comprehension Powered (MCP) technology to convert natural language system descriptions into audit-ready artifacts in OSCAL format for continuous automated compliance…
Can Agents Generalize to the Open World? Unveiling the Fragility of Static Training in Tool Use
Researchers introduce OpenAgent, a problem setting designed to help agents generalize better in real-world scenarios with dynamic queries, tools, and interactions. The study reveals that both Supervised Fine-Tuning and…
MCP Server Architecture Patterns for LLM-Integrated Applications
This industry experience paper presents five architectural patterns observed in the use of Model Context Protocol (MCP) servers, which connect large language models to external tools, data sources, and services. The…
ShareLock: A Stealthy Multi-Tool Threshold Poisoning Attack Against MCP
The paper introduces ShareLock, a stealthy multi-tool threshold poisoning framework designed to exploit the interaction between Large Language Models (LLMs) and external tools through the Model Context Protocol (MCP)…
ShareLock: A Stealthy Multi-Tool Threshold Poisoning Attack Against MCP
The paper introduces a new multi-tool threshold poisoning framework called ShareLock. This framework uses Shamir's threshold scheme to distribute malicious instructions across multiple tool descriptions, making it hard…
Introducing Web Search on Amazon Bedrock AgentCore
AWS announced general availability of web search capabilities for Bedrock agents, with simplified integration for agent tool use.
