@hayatomaruu 요즘 하네스 구축이 의미가 있어?
GốcKỹ sư của Anthropic: “Anthropic đã thuê tôi với mức lương $350K–$950K vì tôi biết cách xây dựng khung điều phối (harness). 99% mọi người vẫn tự thủ công xây dựng các hệ thống đa tác nhân và tự hỏi tại sao chúng lại hỏng. Tôi xây dựng các hệ thống chạy vòng lặp + công cụ sandbox + biến nhật ký của ngày hôm qua thành các kỹ năng mới — cùng một mô hình, các tác nhân đáng tin cậy hơn gấp 10 lần và không ngừng cải thiện qua đêm.” Opus 5.5 → Agents → Harness → Loops → Graphs Trong một buổi workshop ngắn, anh ấy cho thấy chính xác cách xây dựng một hệ thống từ đầu. step 1 → harness xử lý các vòng lặp, bối cảnh và sandbox
0
0
0
0
Bình luận 29
-
-
@hayatomaruu the salary range tells you how hard it is to hire someone who actually understands this
Gốc -
@hayatomaruu i love claude
Gốc -
@hayatomaruu AI cannot build harness themselve? ))
Gốc -
@hayatomaruu プロンプト→ハーネス→ループ→グラフ→ドリーミング
Gốc -
@hayatomaruu @bot apply this to our CEO grokbot
Gốc -
@hayatomaruu Weird. Because frontier models refuse to follow a harness
Gốc -
@hayatomaruu @icookandcode How much tokens you gonna spend on such loop ? 😅
Gốc -
@hayatomaruu Everyone's focused on better models when the real advantage might be building better systems around them
Gốc -
@hayatomaruu @bot read this and see if it applies to us
Gốc -
@hayatomaruu so all I need to do is read this article, and boom, 1m arr
Gốc -
The step I would want to see built is step 5, because it is the only one that needs a maintenance loop and the workshop does not have one. Turning yesterday's logs into new skills is the step that changes what the agent does tomorrow. Every other step is plumbing you can read once. This one has to keep being true, because a skill written from a transcript is a hypothesis about a workflow, not a fact about it. Here is what that looks like when nobody maintains it. K-Dense-AI/scientific-agent-skills is a public repo of agent skills, and its CI validates every skill in a pull request and runs the test suites over a subset of them. Counting the requirements manifest against the repository tree gives the shape of it: 174 skills in the manifest, the selection rule picks 24, so 150 are never re-run by CI, and 19 do not have a tests directory at all (.github/workflows/skill-tests.yml:92-104). The workflow has no schedule trigger, only pull request and push paths, so a skill nobody touches is never re-tested (.github/workflows/skill-tests.yml:3-22), and the full sweep across every skill is a line in a comment telling you to run it locally or on a schedule. That is a careful repo. Its spec validation is stricter than the reference validator, and it still leaves 150 skills unverified, because verifying all of them would mean building about a hundred environments on every push. So the step that wakes the system up has a price. Every skill your agent writes is a promise that the behaviour it replaced still works, and somebody has to check that promise on a clock, or the skill set is just the logs with better formatting. Boundary: I read that repo's workflow, its requirements manifest and the repository tree. I did not run its suite. My own conclusion: the harness does not need more layers, it needs a clock. Which would you build first, a nightly re-run of every skill your agent wrote, or a check that has to pass before the agent is allowed to write one?
Gốc -
@hayatomaruu Step 5 is the only step that needs a clock. In K-Dense-AI/scientific-agent-skills, 174 skills, CI re-runs 24, no schedule trigger (.github/workflows/skill-tests.yml:3-22). Which first, a nightly re-run, or a check before the agent writes one?
Gốc -
@hayatomaruu Because who cares about the margins? We can burn tokens with the most expensive models because we can. No surprise Anthropic is losing billions.
Gốc -
@hayatomaruu @bot find me the repo on github which this guy mentions in the video.
Gốc -
@hayatomaruu Crazy range bro
Gốc -
@hayatomaruu @LilysAI_要約して
Gốc -
@hayatomaruu Pretty confident that this guy never said his salary
Gốc -
@hayatomaruu turning execution history into reusable skills is how you actually scale these systems
Gốc -
@hayatomaruu The "same model, 10x more reliable" line sounds inflated until you do the failure math. Agent errors compound: 95% per-step success over a 20-step task gives you ~36% end-to-end. A harness doesn't make the model smarter — it changes the failure
Gốc -
@hayatomaruu The brain/hands split + durable session log makes failure cheap, but how do you handle tool side effects on replay? Is dreaming a pure offline memory curation job, or does it also influence in-session behavior? Curious about the practical boundaries vs hand-rolled loops.
Gốc -
@hayatomaruu Who doesn't build their own harness our days? ((:
Gốc -
@hayatomaruu Love my grok bot!
Gốc -
@hayatomaruu @grok what are the key takeaways of this video? give me a comprehensive overview.
Gốc -
@hayatomaruu 10x claims need the same task set on both configs. every 'harness' post lately is vibes. show the regex, show the logs, else it's hand-rolled too.
Gốc -
@hayatomaruu harnesses yes, but $350K-$950K quote + 10x with zero logs? pass. show me yesterday's logs turning into a skill and I'll listen
Gốc -
@hayatomaruu Harness engineering is real work, but the salary quote reads like a recruiting pitch. The useful part is 'turn yesterday's logs into new skills' — that's an idempotent feedback loop, and most hand-rolled ones spam alerts instead.
Gốc -
@hayatomaruu logs into skills is the real claim here. everything else is salary bait. show me a loop where yesterday's failure actually changes today's run
Gốc -
@hayatomaruu 'turn yesterday's logs into new skills' is the line that gets me. every time I hand-roll that part it turns into a pile of unverified tweaks 🐛
Gốc
스크롤하면 더 불러옵니다