LACUNA: Safe Agents as Recursive Program Holes

• URL: http://arxiv.org/abs/2605.28617v1
• Date: Wed, 27 May 2026 15:27:25 GMT
• Title: LACUNA: Safe Agents as Recursive Program Holes
• Abstract summary: LLM agents increasingly act by writing code, yet a split persists between the runtime that drives the agent and the code the model writes.<n>We present LACUNA, a programming model for agents that closes this split while preserving safety.<n>Our primitive expresses ReAct loops, sub-agents, skills, parallel decomposition, and multi-model planning as ordinary control flow.
• Score: 3.2613419151327343
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: LLM agents increasingly act by writing code, yet a split persists between the runtime that drives the agent and the code the model writes. The runtime owns the loop, context, and control flow, and the model has little say over any of them. Letting model-written code shape the runtime itself would make agents more expressive, but it would also sharpen safety problems. A model can be diverted by a prompt injection, call the wrong tool, or fail partway and leave an inconsistent state, and each such failure reaches further when the code shapes the runtime than when it expresses a single action. We present LACUNA, a programming model for agents that closes this split while preserving safety. Each agent action is a typed call $\texttt{agent[T](task)}$ that the LLM fills with code when execution reaches it, and the code is type-checked against the surrounding program before it runs. Because each action is accepted or rejected as a whole, a rejected one leaves the environment untouched, and its compiler diagnostics drive a retry. The same check also bounds which tools and data an action may use and how they flow. Our primitive expresses ReAct loops, sub-agents, skills, parallel decomposition, and multi-model planning as ordinary control flow. We evaluate LACUNA on a collection of test cases, BrowseComp-Plus, and $τ^2$-bench. On BrowseComp-Plus, $8.6\%$ of generations are rejected before execution, with 0.7 retries per query on average, and the agent reaches $27.1\%$ accuracy. On $τ^2$-bench, LACUNA solves $76.0\%$ of $392$ tasks across four domains with a capable model, on par with the baseline agent.

Related papers

• Codebook Agent: Amortized Topology Design for LLM Multi-Agent Systems [67.04448659688579]
  We develop a query-independent 16-entry codebook that ranks the top decoded candidates in a single batched forward pass.<n>With no iterative search and no message passing at test time, Codebook Agent is the most accurate method on all six benchmarks.
  arXiv  Detail & Related papers  (2026-09-02T08:10:22Z)
• LoopArena: Benchmarking Models as Runtime Controllers for Loop Engineering [47.39676953516343]
  Loop Engineering is emerging as a practice for organizing development work around coding agents.<n>We introduce LoopArena, a benchmark for evaluating how well one model can guide a separate coding agent through a long-running task.
  arXiv  Detail & Related papers  (2026-08-28T12:44:54Z)
• Long-Horizon State Tracking in LLMs: Executing MD5 through a Deep Sequence of Dependent Tool Calls [0.5097809301149342]
  Long-horizon tasks remain uncommon in large language model (LLM) evaluation.<n>We test whether an LLM can carry exact intermediate state across many tool calls at all.<n>Two ingredients decide success: keeping the model's own reasoning in its context each turn, and voting over a thinking-enabled worker to remove its modular-arithmetic slips.
  arXiv  Detail & Related papers  (2026-08-02T09:38:02Z)
• When Does Restricting a Coding Agent to execute_code Help? A Regime $\ imes$ Agent-Design Ablation [20.672156968511587]
  We run a comparison of three-arm ablation (baseline / bash_only / code_only) on synthetic computation tasks and SWE-bench Mini modification tasks.<n>Across the four resulting (regime, agent) cells, restricting the agent to a single execute_code MCP tool is cheaper than -- or statistically tied with -- its cheapest tool-rich rival.<n>The lone exception is SWE-bench/Claude, where code_only is directionally costlier.
  arXiv  Detail & Related papers  (2026-07-12T04:52:08Z)
• Confining Nondeterminism: AI-Driven Research Systems as DBMSs for Reliable, Non-Wasteful, Transparent, and Collaborative Research [Vision] [6.1118660156378075]
  We argue that today's agent loop is an executor whose internal state nobody, including the agent, can check.<n>Rather than trying to see inside the LLM, we take a lesson from databases, which earn trust without being watched.<n>The project lives in a deterministic, versioned dataflow engine, and the LLM, together with the user, is a compiler that may only edit that plan.
  arXiv  Detail & Related papers  (2026-07-11T23:46:01Z)
• AgentLens: Production-Assessed Trajectory Reviews for Coding Agent Evaluation [0.18567307896771823]
  AgentLens is a production-assessed benchmark for interactive code agents.<n>It pairs formal verification with LLM-written trajectory reviews and side-by-side comparisons.<n>We use it to diagnose model behavior, compare successive versions of our own agent, and catch product regressions in a nightly evaluation pipeline.
  arXiv  Detail & Related papers  (2026-07-07T11:27:43Z)
• 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)
• CoAgent: Concurrency Control for Multi-Agent Systems [3.127039011380953]
  This paper builds control on a capability classical transactions lack.<n>It fixes a serialization order at launch, serves each read the order-filtered value, and applies writes speculatively in place.<n>It grows a 25-tool library online and lifts the task pass rate from 45/71 to 63/71 at $0.80times$ the time and $0.86times$ the cost.
  arXiv  Detail & Related papers  (2026-06-13T16:15:14Z)
• 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)
• ISO-Bench: Can Coding Agents Optimize Real-World Inference Workloads? [0.8749675983608171]
  We introduce ISO-Bench, a benchmark for coding agents to test their capabilities on real-world inference tasks.<n>We curated 54 tasks from merged pull requests with measurable performance improvements.
  arXiv  Detail & Related papers  (2026-02-23T08:37:53Z)
• AgentArk: Distilling Multi-Agent Intelligence into a Single LLM Agent [57.10083973844841]
  AgentArk is a novel framework to distill multi-agent dynamics into the weights of a single model.<n>We investigate three hierarchical distillation strategies across various models, tasks, scaling, and scenarios.<n>By shifting the burden of computation from inference to training, the distilled models preserve the efficiency of one agent while exhibiting strong reasoning and self-correction performance of multiple agents.
  arXiv  Detail & Related papers  (2026-02-03T19:18:28Z)
• CodeMem: Architecting Reproducible Agents via Dynamic MCP and Procedural Memory [0.5599792629509229]
  Current tool-using AI agents suffer from limited action space, context inefficiency, and probabilistic instability.<n>This paper proposes CodeMem, an architecture to implement procedural memory via code which can be used to build and run reusable agentic with deterministic reliability.
  arXiv  Detail & Related papers  (2025-12-17T11:28:25Z)
• Holistic Agent Leaderboard: The Missing Infrastructure for AI Agent Evaluation [87.47155146067962]
  We provide a standardized evaluation harness that orchestrates parallel evaluations across hundreds of tasks.<n>We conduct three-dimensional analysis spanning models, scaffolds, and benchmarks.<n>Our analysis reveals surprising insights, such as higher reasoning effort reducing accuracy in the majority of runs.
  arXiv  Detail & Related papers  (2025-10-13T22:22:28Z)
• Runaway is Ashamed, But Helpful: On the Early-Exit Behavior of Large Language Model-based Agents in Embodied Environments [54.67512489842682]
  Large language models (LLMs) have demonstrated strong planning and decision-making capabilities in complex embodied environments.<n>We take a first step toward exploring the early-exit behavior for LLM-based agents.
  arXiv  Detail & Related papers  (2025-05-23T08:23:36Z)
• CodeCoR: An LLM-Based Self-Reflective Multi-Agent Framework for Code Generation [10.048098631259876]
  Code generation aims to produce code that fulfills requirements written in natural languages automatically.<n>Large language Models (LLMs) like ChatGPT fail to ensure the syntactic and semantic correctness of the generated code.<n>We propose CodeCoR, a self-reflective multi-agent framework that evaluates the effectiveness of each agent and their collaborations.
  arXiv  Detail & Related papers  (2025-01-14T03:21:10Z)

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.