LACUNA: Safe Agents as Recursive Program Holes
:::info Stub — Full Engineering Breakdown Coming This paper was featured on Hugging Face Daily Papers on 2026-05-27 with 4 upvotes. A full breakdown with production viability rating, implementation notes, and honest limitations is being written. Subscribe to AI Letters → :::
| Authors | Yaoyu Zhao et al. |
| Year | 2026 |
| HF Upvotes | 4 |
| arXiv | 2605.28617 |
| Download | |
| HF Page | View on Hugging Face |
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 agentT 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.
Engineering Breakdown
The Problem
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.
The Approach
We present LACUNA, a programming model for agents that closes this split while preserving safety.
Key Results
On τ^2-bench, LACUNA solves 76.0% of 392 tasks across four domains with a capable model, on par with the baseline agent.
Research Areas
This paper contributes to the following areas of AI/ML engineering:
- Machine learning
- Deep learning
- Neural networks
- Model optimization
- AI systems
- Recursive
:::tip Subscribe Get weekly breakdowns of papers like this in AI Letters - the newsletter for engineers building production AI systems. :::