DynamicMCPBench: A Trace-Grounded, Effect-Scored Benchmark for LLM Agents over Live MCP Servers

• URL: http://arxiv.org/abs/2607.20531v1
• Date: Fri, 10 Jul 2026 12:25:57 GMT
• Title: DynamicMCPBench: A Trace-Grounded, Effect-Scored Benchmark for LLM Agents over Live MCP Servers
• Abstract summary: We present DynamicMCPBench, a reusable framework rather than a fixed dataset.<n>It generates realistic goals, pursues each one live to record a successful trajectory, distills that trajectory into path-agnostic effect checkpoints, and scores an agent on whether it reproduces those effects, never on the final answer.<n>Even the strongest agents solve only about half of the tasks, 31% of tasks are solved by no model at all, and accuracy collapses as the required tool chain grows longer.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language model (LLM) agents are increasingly deployed over Model Context Protocol (MCP) servers, yet the benchmarks used to evaluate them score the final answer or a fixed "ground-truth" list of tools, both of which are fragile once the underlying data is live and stateful. We present DynamicMCPBench, a reusable framework rather than a fixed dataset. A practitioner can run it on their own MCP servers to test models on their own tasks, or let it collect servers automatically to measure a model's general ability to solve agentic tasks. Given the servers and any set of models, it generates realistic goals, pursues each one live to record a successful trajectory, distills that trajectory into path-agnostic effect checkpoints, and scores an agent on whether it reproduces those effects, never on the final answer. To show what the framework reveals, we run it at scale: 24 models over 121 servers and 750 tasks spread evenly over 15 task categories (50 each), where each category targets a distinct tool-use challenge of the generated questions. Each task is scored by pass^3: it counts as solved only if all three independent attempts succeed. Even the strongest agents solve only about half of the tasks, 31% of tasks are solved by no model at all, and accuracy collapses as the required tool chain grows longer (from 39% on the shortest chains to 13% on the longest). A human validation study confirms the automatic scoring is reliable (chance-corrected agreement of 0.76). DynamicMCPBench thus turns benchmark construction into something practitioners can rerun on their own servers and models, while exposing a consistent inability of current agents to handle long, multi-step agentic tasks.

Related papers

• BekchiAI: Measuring, Observing, and Controlling LLM Agents in One Click [0.0]
  We present BekchiAI, a benchmark for measuring agentic skill and a platform for observing and controlling live agents.<n>The BekchiAI-Platform is a complementary web-based observability and control layer for deployed agents, providing full token and telemetry as well as remote run termination.
  arXiv  Detail & Related papers  (2026-08-27T09:30:08Z)
• Joint Optimization of Tool Creation and Use for Large Language Model Agents [69.07066297088187]
  SMITH (-grounded Multi-task Iterative Tool Honing) trains tool creation and tool use inside a single policy.<n>A 4B Qwen3 trained with SMITH on 13 procedural reasoning tasks with exact verifiers reaches 79.8 macro-average accuracy on held-out tasks.<n>It also reaches 40.4 on TabMWP-Hard and 42.6 on out-of-domain GQA (+7.6 over the best same-backbone inference-time baseline)
  arXiv  Detail & Related papers  (2026-08-25T13:59:34Z)
• ClawArena-Team: Benchmarking Subagent Orchestration and Dynamic Workflows in Language-Model Agents [61.184954205350984]
  ClawArena-Team is a benchmark of 41 multi-turn, multimodal, multi-agent scenarios spanning 258 evaluation rounds and 72 staged updates.<n>All scoring is execution-based with no LLM judge: an overall score -- the Subagent-Management Score (SMS) -- multiplies task correctness by a least-privilege and modality-routing factor.
  arXiv  Detail & Related papers  (2026-06-30T06:06:08Z)
• GTA: Generating Long-Horizon Tasks for Web Agents at Scale [82.43869456830664]
  We introduce a scalable framework, GTA, that integrates crawling, retrieval-based seeding, in-context generation, and automated quality control.<n>We instantiate the pipeline on over 50 websites covering e-commerce, government, forums, and news, with multilingual and multi-hop coverage.<n>Our contributions are three-fold: (i) formalizing multi-hop web-agent task generation, (ii) proposing an efficient and validated pipeline for automatic data creation, and (iii) releasing a dynamic benchmark with reproducible evaluation.
  arXiv  Detail & Related papers  (2026-05-28T01:05:50Z)
• AgentAtlas: Beyond Outcome Leaderboards for LLM Agents [0.025718125188898048]
  AgentAtlas reframes agent evaluation as a diagnostic vocabulary and audit protocol.<n>The paper contributes: (i) a six-state control-decision taxonomy (Act / Ask / Refuse / Stop / Confirm / Recover); (ii) a trajectory-failure vocabulary with primary error source and downstream impact; and (iv) an illustrative protocol study on a synthetic 1,342-item set evaluated with eight models.
  arXiv  Detail & Related papers  (2026-05-19T22:05:12Z)
• EnactToM: An Evolving Benchmark for Functional Theory of Mind in Embodied Agents [75.01735520608075]
  Existing benchmarks mostly test literal Theory of Mind (ToM) by asking direct belief questions.<n>We introduce EnactToM, an evolving benchmark of 300 embodied multi-agent tasks set in a 3D household.<n>On the hard split, all seven evaluated frontier models score 0.0% Pass3 on functional task completion, while averaging 45.0% on literal belief probes.
  arXiv  Detail & Related papers  (2026-05-11T00:04:19Z)
• Towards Multi-Agent Autonomous Reasoning in Hydrodynamics [0.06999740786886537]
  We present a multi-agent system (MAS) prototype for hydrodynamics in which specialized agents are coordinated through a Layer Execution Graph (LEG)<n>A planner agent constructs query-specific execution topologies from natural-language routings that capture domain knowledge without hard-coding it as rigid control logic.<n>A reporter agent synthesizes the final response, while the runtime logs provenance for every tool invocation to support auditability.
  arXiv  Detail & Related papers  (2026-05-01T21:17:55Z)
• ClawMark: A Living-World Benchmark for Multi-Turn, Multi-Day, Multimodal Coworker Agents [77.22389710754452]
  We introduce bench, a benchmark for coworker agents built around multi-turn multi-day tasks.<n>The current release contains 100 tasks across 13 professional scenarios, executed against five stateful sandboxed services.<n>The strongest model reaches 75.8 weighted score, but the best strict Task Success is only 20.0%, indicating that partial progress is common.
  arXiv  Detail & Related papers  (2026-04-26T16:05:02Z)
• Agentic-MME: What Agentic Capability Really Brings to Multimodal Intelligence? [35.30497528897595]
  Agentic-MME is a process-verified benchmark for Multimodal Agentic Capabilities.<n>It contains 418 real-world tasks across 6 domains and 3 difficulty levels.<n>It features over 2,000 stepwise checkpoints that average 10+ person-hours of manual annotation per task.
  arXiv  Detail & Related papers  (2026-04-03T13:02:01Z)
• RewardHackingAgents: Benchmarking Evaluation Integrity for LLM ML-Engineering Agents [0.9821874476902969]
  LLM agents increasingly perform end-to-end ML engineering tasks where success is judged by a single scalar test metric.<n>This creates a structural vulnerability: an agent can increase the reported score by compromising the evaluation pipeline rather than improving the model.<n>We introduce RewardHackingAgents, a workspace-based benchmark that makes two compromise vectors explicit and measurable.
  arXiv  Detail & Related papers  (2026-03-11T22:06:44Z)
• Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systems [50.29939179830491]
  Failure attribution in LLM multi-agent systems remains underexplored and labor-intensive.<n>We develop and evaluate three automated failure attribution methods, summarizing their corresponding pros and cons.<n>The best method achieves 53.5% accuracy in identifying failure-responsible agents but only 14.2% in pinpointing failure steps.
  arXiv  Detail & Related papers  (2025-04-30T23:09:44Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.