The paper discusses how large enterprises can adopt the harness paradigm to overcome limitations of traditional coding approaches. It reviews recent findings that harnesses outperform complex architectures at the task level, that harness choice drives benchmark variance more than model choice, and that governance is the main barrier to enterprise adoption. The authors propose a unified harness architecture that keeps code identical across deployments, simplifying review and audit processes.
arXiv:2607. 06906v1 Announce Type: new Abstract: Agentic AI development today runs on token maxing: buying capability with tokens -- longer reasoning traces, more turns, wider tool payloads, bigger replayed contexts -- so tokens per task grow faster than task value.
arXiv:2606. 11869v1 Announce Type: cross Abstract: Custom AI agents areagents that live inside their own application, talk to their own data and tools, enforce their own security boundaries, and carry their own brand and audit trail.
Agentic AI development today runs on token maxing: buying capability with tokens -- longer reasoning traces, more turns, wider tool payloads, bigger replayed contexts -- so tokens per task grow faster than task value. Falling per-token prices mask the pattern; total spend rises anyway.
The paper examines three open-source agent harnesses—LangChain’s deepagents, Earendil’s pi, and DeepSeek’s dsh—each built from contrasting design philosophies. By analyzing their source code and commit histories, the authors find that the mature harnesses converge on five common architectural elements: a commoditized loop, an append‑only replayable session record, model quirks stored as data, progressive disclosure of context, and explicit extension seams. A fourth harness, used as a held‑out check, also displays all five elements and even reuses another’s implementation, indicating that convergence arises from parallel discovery, diffusion, and literal reuse rather than independent invention. The study notes a missing dimension—external verifiability via a tamper‑evident record—highlighting a future axis for provenance‑sensitive domains.
Each new generation of frontier model pushes the edge of what is possible. Inside a harness -- a loop, a shell, a filesystem, the ability to write and run its own code -- a model can take on increasin...
arXiv:2606. 10106v1 Announce Type: cross Abstract: The term agent harness now circulates widely in software engineering with generative artificial intelligence.
The paper defines superintelligent software engineering as building and evolving software with agents whose capability exceeds that of human engineers, emphasizing that governance must scale with capability. It argues that correctness depends on decisions and contracts beyond code, that knowledge decays, and that when agents outpace reviewers, stale knowledge grows. The authors introduce OSTEON, a governance framework that tracks decisions, contracts, and knowledge decay, and report no measured effect of the system while outlining questions that could falsify their thesis.
The paper introduces Growing Harness, a training method that transforms recurring control logic in large language model agents into reusable executable code, reducing reliance on the model for task-specific decisions. By using strategy-free scaffolds, failure-guided code repair, and success-first gating, the approach learns a shared harness that improves performance across multiple benchmarks and model sizes. Experiments on BrowseComp-Plus and WebArena-Verified show significant gains in success rates and substantial reductions in LLM calls and inference cost compared to traditional tool‑calling agents.
An agent harness is what turns a language model into an autonomous agent: the surrounding code that builds the model's context, mediates its tools, runs the loop, and persists state across a long-hori...
arXiv:2607. 08028v1 Announce Type: new Abstract: Enterprise large language model (LLM) applications often begin as prototypes whose behavior is carried by prompts and retrieval context.
The paper introduces kw-bench, a benchmark that evaluates coding-agent harnesses on investigative knowledge work, specifically a five‑year VAT reconstruction task across multiple SAP systems. It contrasts generative benchmarks with this real‑world, evidence‑heavy task, showing that model‑agnostic harnesses performed well while lab harnesses failed. The study demonstrates how frontier models can dramatically reduce time and cost for complex audit‑style engagements.
The paper introduces agent‑native primitives—reusable building blocks that embed a language model within a resource to apply it contextually—and HEELO, a harness‑engineering framework that composes these primitives into task‑specific graphs, verifies outcomes, and iteratively refines or adds primitives. Using eight primitive families, the authors demonstrate significant performance gains across multiple benchmarks, raising Terminal‑Bench 4.0 resolution from 37.3% to 55.5% and improving pass rates on ToolBench and GAIA‑Text‑103. When integrated with HEELO, Claude Opus 5.5 and GPT‑6.1 Sol achieve higher completion counts, faster runtimes, and lower costs on 46 real‑world tasks.