OSFoundry: Building and Evolving Operating Systems with Specification-Guided Agents
On the map: LLM agents and tool use → Software engineering coding agents
Abstract (arXiv)
Operating systems must evolve continuously. Yet their development remains code-centric and largely manual: even a localized change can require recovering implicit assumptions, coordinating multiple subsystems, and repeatedly building, booting, testing, and debugging the complete system. General-purpose coding agents automate individual edits, but their prompt-centric workflows repeatedly reconstruct task boundaries and OS semantics from scattered context, limiting their reliability for sustained OS evolution. This paper presents OSFoundry, an OS-specialized agent harness that makes design intent persistent throughout OS development. Its key insight is to separate stable intent from diverse implementation. Instead of relying on free-form natural-language prompts to convey design intent, OSFoundry uses SysSpec*, a shared development blueprint comprising a task-bounding Plan and an OS-specific Specification: the Plan bounds what a change should achieve, while the Specification records the interfaces, modular dependencies, and concurrency semantics that its implementation must preserve. Agents implement and validate code against this same blueprint, refining both SysSpec* and the implementation when execution or review exposes a mismatch. We evaluate OSFoundry along three ways. First, OSFoundry generates SpecOS from SysSpec*; the resulting complete OS boots and passes all 70 functional tests. Second, specification patches evolve SpecOS with a GUI and three performance optimizations that improve performance by up to 4.41x. Third, across 11 recent Linux-kernel bug-fix tasks, OSFoundry achieves 1.8x the accuracy of Codex using GPT-5.5. These results show that persistent specifications can shift OS construction and evolution from repeated manual kernel engineering toward specification-guided development.
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use. Builds on 27 of its 53 references that are in the map. First posted on arXiv 8 Aug 2026.
Builds on
27 of 53 references.
Closest ancestors · 12
Same direction, newest first
- 2026DeepSWE: Measuring Frontier Coding Agents on Original, Long-Horizon Engineering Tasks8 Jul
- 2026VibeServe: Can AI Agents Build Bespoke LLM Serving Systems?7 May
- 2026Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses28 Apr
- 2026CUDA Agent: Large-Scale Agentic RL for High-Performance CUDA Kernel Generation27 Feb
- 2026VeRO: A Harness for Agents to Optimize Agents25 Feb
- 2025Code Researcher: Deep Research Agent for Large Systems Code and Commit History27 May
- 2025Taming System Complexity: Demystifying Software Engineering Agents in Diagnosing Linux Kernel Faults26 May
- 2025kAgent: An execution-guided crash resolution agent for the Linux kernel29 Apr
- 2024Agentless: Demystifying LLM-based Software Engineering Agents1 Jul
- 2024SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering6 May
- 2024RepairAgent: An Autonomous, LLM-Based Agent for Program Repair25 Mar
- 2024CodeAgent: Enhancing Code Generation with Tool-Integrated Agent Systems for Real-World Repo-level Coding Challenges14 Jan
Other references · 15
- 2024AutoCodeRover: Autonomous Program Improvementcited by 272
- 2025SWE-Bench Pro: Can AI Agents Solve Long-Horizon Software Engineering Tasks?cited by 254
- 2023Clover: Closed-Loop Verifiable Code Generationcited by 62
- 2024SpecGen: Automated Generation of Formal Program Specifications via Large Language Modelscited by 51
- 2026ProgramBench: Can Language Models Rebuild Programs From Scratch?cited by 47
- 2024AutoVerus: Automated Proof Generation for Rust Codecited by 46
- 2024AlphaVerus: Bootstrapping Formally Verified Code Generation through Self-Improving Translation and Treefinementcited by 33
- 2024Automated Proof Generation for Rust Code via Self-Evolutioncited by 31
- 2023KernelGPT: Enhanced Kernel Fuzzing via Large Language Modelscited by 17
- 2026Spec-Driven Development:From Code to Contract in the Age of AI Coding Assistantscited by 14
- 2025OSVBench: Benchmarking LLMs on Specification Generation Tasks for Operating System Verificationcited by 8
- 2026SEC-bench Pro: Can Language Models Solve Long-Horizon Software Security Tasks?cited by 8
3 more
- 2025QiMeng-Xpiler: Transcompiling Tensor Programs for Deep Learning Systems with a Neural-Symbolic Approachcited by 4
- 2025Sharpen the Spec, Cut the Code: A Case for Generative File System with SYSSPECcited by 4
- 2026SpecDB: LLM-Generated Customized Databases via Feature-Oriented Decompositioncited by 1
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026AI-Generated Smells: An Analysis of Code and Architecture in LLM and Agent-Driven Development
- 2026Persistent Recursive Worlds Enable Autonomous Software Evolution
- 2026Agent Operating Systems (AOS): Integrating Agentic Control Planes into, and Beyond, Traditional Operating Systems
- 2026Towards an Agent Operating System - Lessons from Classical and Cloud OS
- 2026SpecFirst: Behavioral Specification Elicitation as a First-Class Step in Agent-Based Program Synthesis from Scratch
- 2026SWE-Milestone: Evaluating AI Agents on Continuous Software Evolution
- 2026Octopus Protocol: One-Shot Hardware Discovery and Control for AI Agents via Infrastructure-as-Prompts
- 2026Needle in the Repo: A Benchmark for Maintainability in AI-Generated Repository Edits