Many agent failures are not only model failures, but harness/system failures. As agents move into production, context, memory, tool routing, orchestration, and verification need to be treated as first-class design and evaluation targets.</p>\n","updatedAt":"2026-06-01T16:01:41.220Z","author":{"_id":"64b78c09964f6e7bf32dd5f3","avatarUrl":"/avatars/e12e82421a7c677aa513b6418487defa.svg","fullname":"Shangding Gu","name":"Shangding-Gu","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":1,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.9250882267951965},"editors":["Shangding-Gu"],"editorAvatarUrls":["/avatars/e12e82421a7c677aa513b6418487defa.svg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2605.26112","authors":[{"_id":"6a1dac2b808ddbc3c7d4398c","name":"Shangding Gu","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/64b78c09964f6e7bf32dd5f3/pJxtkW1_tGedtQ0zAbOF3.gif"],"publishedAt":"2026-05-25T00:00:00.000Z","submittedOnDailyAt":"2026-06-01T00:00:00.000Z","title":"From Model Scaling to System Scaling: Scaling the Harness in Agentic AI","submittedOnDailyBy":{"_id":"64b78c09964f6e7bf32dd5f3","avatarUrl":"/avatars/e12e82421a7c677aa513b6418487defa.svg","isPro":false,"fullname":"Shangding Gu","user":"Shangding-Gu","type":"user","name":"Shangding-Gu"},"summary":"This paper studies the next major bottleneck in agentic AI as system scaling, not only model scaling: the design of auditable, persistent, modular, and verifiable architectures around foundation models. We refer to this shift as scaling the harness: treating the structured execution layer around a foundation model as a first-class object of design, evaluation, and optimization. Although recent large language models enable agents to use tools, retrieve information, maintain memory, and execute long-horizon workflows, evaluation remains largely model-centric, often reducing agents to final-task success while treating memory, retrieval, tool use, orchestration, verification, and governance as secondary implementation details. This framing is increasingly inadequate because agent performance emerges from the interaction among the foundation model, memory substrate, context constructor, skill-routing layer, orchestration loop, and verification-and-governance layer. Together, these components form the agent harness, which translates model capability into long-horizon agent behavior. We study scaling the harness through three core bottlenecks: context governance, trustworthy memory, and dynamic skill routing, together with the orchestration and governance mechanisms that coordinate and constrain them. We further outline a research agenda for harness-level benchmarks that go beyond one-shot task success to measure trajectory quality, memory hygiene, context efficiency, communication fidelity, verification cost, and safe evolution over time. To make the discussion concrete, we develop CheetahClaws: https://github.com/SafeRL-Lab/cheetahclaws, a Python-native reference harness, and compare it with Claude Code and OpenClaw. Our main claim is that future progress in agentic AI will depend as much on system design as on stronger foundation models.","upvotes":1,"discussionId":"6a1dac2b808ddbc3c7d4398d","projectPage":"https://cheetahclaws.github.io/","githubRepo":"https://github.com/SafeRL-Lab/cheetahclaws","githubRepoAddedBy":"user","ai_summary":"Agentic AI advancement requires scaling system architecture around foundation models, focusing on auditable and verifiable components rather than just model capacity.","ai_keywords":["agentic AI","foundation models","system scaling","auditable architectures","persistent architectures","modular architectures","verifiable architectures","scaling the harness","structured execution layer","agent harness","memory substrate","context constructor","skill-routing layer","orchestration loop","verification-and-governance layer","context governance","trustworthy memory","dynamic skill routing","harness-level benchmarks","trajectory quality","memory hygiene","context efficiency","communication fidelity","verification cost","safe evolution"],"githubStars":711,"organization":{"_id":"61f20a9ce108f2cba2dc0730","name":"Berkeley","fullname":"UC Berkeley","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/61ac8f8a00d01045fca0ad2f/0FjsTg2txEZZ4dEgmMnQL.png"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"64b78c09964f6e7bf32dd5f3","avatarUrl":"/avatars/e12e82421a7c677aa513b6418487defa.svg","isPro":false,"fullname":"Shangding Gu","user":"Shangding-Gu","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"61f20a9ce108f2cba2dc0730","name":"Berkeley","fullname":"UC Berkeley","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/61ac8f8a00d01045fca0ad2f/0FjsTg2txEZZ4dEgmMnQL.png"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2605/2605.26112.md"}">

Get this paper in your agent:

Don't have the latest CLI?

Cite arxiv.org/abs/2605.26112 in a model README.md to link it from this page.

Cite arxiv.org/abs/2605.26112 in a dataset README.md to link it from this page.

Cite arxiv.org/abs/2605.26112 in a Space README.md to link it from this page.

Add this paper to a collection to link it from this page.