Swanky Hsiao | swanbear.eth

@CloneGirls in Taipei! 😍 #CloneGirls #CloneX #RTFKT

Kimi K3 vs GPT-5.6 Sol vs Fable 5 Three top-tier models. Same /design prompt across all. Parameters reviewed: Pace, design sense, gameplay feel Result: Kimi K3 nails the design sense and knows exactly what to add in-game. Pace and gameplay held up well. Fable 5 and GPT-5.6 Sol both play too fast, video isn't sped up. Ranking (DX, Features & Cost): → Kimi K3: 9.5/10 · $0.030 → Fable 5: 7.5/10 · $0.38 → GPT-5.6 Sol: 7/10 · $0.11

I talk to engineers at other companies every day and hear the same thing: one person is 10x'ing their output with Claude but the rest of the org hasn't caught up. Watching teams adopt AI, I keep seeing the same 4 steps. I mapped them out here: Steps of AI Adoption claude.ai/code/artifact/…

You can now orchestrate Fable 5, Sol, and any model inside Codex with one plugin. It's called Codex-Orchestration. Assign Fable 5 as the advisor, Sol as the executor, or any model to any role. Then define the order they work in. Codex handles the routing. I ran Fable 5 High as planner with GPT-5.6 Sol Extra High as executor on a set of issues Opus and GPT-5.5 always struggled with. Done in 30 minutes. 40% fewer limit hits. 2x faster implementation. Install it by pasting this into Codex: "Install Codex Orchestration:  codex plugin marketplace add Cjbuilds/Codex-Orchestration codex plugin add codex-orchestration@codex-orchestration  Verify the installation, then tell me to start a new task." Then assign your models:  @ codex-orchestration advisor:  Claude Fable 5 High, Executor: GPT-5.6 Sol High Open source. Tweak the routing however you want.

I asked k3: "I want you to push the boundaries of what's possible in web design with the absolute latest in css, webgl, shaders, etc as of July 2026 and create the most impressive, informational single-page experience you can imagine" Result:

最近 Hermes Agent 非常紅，熱度上已經超越了俗稱「龍蝦」的 OpenClaw 我自己玩了一個多月的 Hermes，整理出 10 個最有感的設定 如果你也想加入「養殖業」，不妨參考本篇 blog.kyomind.tw/hermes-agent/

Something I have been thinking about: in the past, the best engineers I knew spent a lot of time automating their work in various ways. Better vim/emacs automations, writing lint rules to catch repeat code issues, building up a suite of e2e tests so they don't need to smoke test the app manually. These kinds of things were the highest leverage activities an engineer could do, because it multiplied their own output, which in turn meant they could build more things. I think many of these automations have become even more important now. This is true for a number of reasons. First, infra and DevX automation speeds you up. And if you are running an army of agents, each of those agents will be sped up also. More automation == more output per unit of time. Second, moving things to code improves efficiency. Your agent could fix an issue every time it sees that issue happen, but that uses tokens and might miss cases. If Claude instead writes a lint rule, CI step, or routine, that class of issue can be fully automated forever. This is really what people are talking about when they talk about loops -- it's about automating entire types of busywork rather than solving them one off. This isn't a new idea at all. Engineers have been doing this for a long time! Third and most importantly, automation makes it possible for others to contribute to the codebase more easily. Increasingly what I am seeing is engineers are contributing to codebases on day one because Claude can navigate the codebase for them, and that non-engineers are able to contribute to a codebase as effectively as engineers can. What gets in the way of both of these is domain knowledge that lives in peoples' heads rather than in automation -- the stuff you used to have to learn when ramping up. What has changed thanks to agents is the domain knowledge that can be encoded as infrastructure is no longer limited to what is expressible in lint rules and types and tests; it can now capture nearly all domain knowledge, encoded as code comments and skills and CLAUDE.md rules and memories. If I put up a PR for an iOS codebase I don't know and a code reviewer rejects it because it doesn't use the right framework, or if a designer builds a new feature and it gets rejected because it doesn't follow the right architectural patterns, these are failures of automation. Every team should be writing the CLAUDE.md's, REVIEW.md's, skills, and docs that enable agents to productively work in their codebase with zero additional context from the prompter. This sounds crazy, and at the same time is a natural extension of the stuff engineers have always done: automate, and encode domain knowledge as infrastructure. As the model gets smarter and as the harness matures, this task becomes easier. In the meantime, it is on every team to look for ways to convert their domain knowledge to infra so that Claude can write code better, so that code review catches issues automatically, and so the next person working on your codebase can contribute more easily.

用 Hermes Agent 串 ChatGPT 生 CloneX 塔羅圖 之前一張一張手動貼 prompt 現在一次生完也自己檢查重做 看來 @ucx_tw 計劃之後有機會重啟了😍 swanky.github.io/tarot/decks/cl…

@realChainDoctor 沒錯，叫他生個美女照都會說違反風俗

有没有人觉得gpt5.6-sol 道德感太强了？

If you aren't yet bold enough to install the Codex app, you can stay in the presence of your orange crab and point it at GPT 5.6 Sol. Takes 5 minutes. Kudos to Theo for explaining one of the ways to get this done. Step 1: Install CLIProxyAPI Step 2: Connect Step 3: Define following alias and enjoy claudex ``` alias claudex='CLAUDE_CODE_SUBAGENT_MODEL=gpt-5.6-sol \ CLAUDE_CODE_ALWAYS_ENABLE_EFFORT=1 \ CLAUDE_CODE_MAX_TOOL_USE_CONCURRENCY=3 \ ENABLE_TOOL_SEARCH=false \ claude --model gpt-5.6-sol' ``` If this gets blocked, I owe you a reset.

1. Delete the docs you create to explain your code 2. Take the tokens you save on updating those docs 3. Spend them on making your code self-explanatory

AI Coding 選型，早就不是「哪個模型分數最高」這麼簡單。 Claude Code、Codex、Cursor、Copilot、OpenCode、Hermes，看似都在幫你寫程式，實際上卻分別站在模型、開發介面、Agent Harness、工作流與企業治理的不同位置。 我整理了一套「AI Coding 九層選型框架」，涵蓋： Harness、模型與 Provider、AGENTS.md、工具整合、權限治理、團隊協作，以及每個 Accepted Task 的真實成本。 排行榜很快會過期，但好的工程判斷方法，不該每次都推倒重來。 完整文章： swanky.github.io/technical/ai-c… #AICoding #AgenticEngineering #SoftwareEngineering #DeveloperTools #AITransformation

New Hermes Master class video by @tonbistudio !

ChatGPT 的 "Usage limit resets" 真的太強大了 這到底要怎麼輸? 我之前怎麼都沒發現這個功能🥰

@catmanyau @1PercentBetterT It's expensive. Hermes Agent need lots of token. 😅 openrouter.ai/apps/hermes-ag…

@swanky @1PercentBetterT grok 4.5 is a surprise choice here. what’s pulling you away from claude for this agent setup?

[颱風天就是該待在家裡燒 Token] 覺得最近 Opus 4.8 真的幻覺現象很嚴重... 🙃 之後 Fable 5 不能用該怎麼辦? 已經訂了 ChatGPT Pro 接 Hermes Agent 來用了 目前體驗真的還不錯，不過之後還是會想用 Grok 4.5 看看

@1PercentBetterT I'm used to Claude Code workflow, haven't tried Codex much yet. Mainly looking for a Hermes Agent setup that isn't tied to one LLM and has good memory/persistence. Planning to test Grok 4.5 soon too.

@swanky why not GPT with codex ?

原本只是想測 GPT-5.6 Sol，最後卻讓 Hermes Agent 串上 OpenRouter，用 Seedance 2.0 Fast 生成影片鏡頭，再交給 FFmpeg、Pillow 與 Edge TTS 完成本機後製。 真正有意思的不是「AI 會生影片」，而是如何把成本核准、冪等、失敗重試與人工驗收都放進工作流。完整過程與 35 秒成品： swanky.github.io/technical/herm… #HermesAgent #OpenRouter #AIVideo

Grok 4.5 不错，在我的链上攻击分析 Agent 上，实测验证了几次，令我满意。之前的版本我是失望的，但 4.5 可以了。 场景：加密货币链上攻击分析，在 Hermes Agent 里做了 @SlowMist_Team 相关 Harness Engineering。 GPT 5.5/5.6，GLM 5.2，Grok 4.5，DeepSeek V4 Pro/GLM 5.1，这是我的满意度梯队（从很好到还行）。为什么没有 Claude 的？已经不是关键必需品了…还一天天搞歧视及封锁… 以模型和 Agents 这样的迭代速度，相关工作真的会越来越没有门槛，门槛以后只会集中在有没有实践上，实践决定了一个人的经验，经验可以影响许多。

這兩天最重要的 AI 新聞，可能不是任何一個新模型。 是 OpenAI 把桌面版的門牌換了：寫程式用的 Codex 升格成新版 ChatGPT 桌面 App 的主體，那個定義了整個 AI 時代的聊天框，改名叫 ChatGPT Classic。 48 小時、三場發布，分工意外地清楚—— 🔹 Grok 4.5：便宜到分數可能不重要的執行器 🔹 Muse Spark 1.1：不砌牆、想當包工頭的多代理編排者 🔹 GPT-5.6：正面對上 Fable 5 的 agentic coding 上限 🔹 Claude Fable 5：判斷力與長時程工作仍是標竿 這一輪比的不再是誰更會回答，而是誰更會把活做完。 我把這波發布潮整理成一篇文章，包括： ✅ 「Fable 5 拿 80.3%」其實有三個口徑——benchmark 的口徑陷阱怎麼避 ✅ Codex 扶正背後的三組數字（用戶量、估值、收入） ✅ 給台灣團隊的 model routing 假設表：什麼任務該用哪個模型 ✅ agent 說「done」為什麼不能全信 模型正在變成可替換的執行元件。真正值得累積的資產是任務資料集、驗收標準、routing 政策，和一套可重複運作的驗證迴路。 當三家公司都在爭著替你把活幹完——你的工作流，準備好驗收了嗎？ 全文 👉 swanky.github.io/technical/ai-m… #AgenticEngineering #ClaudeCode #GPT56 #AIAgent #ModelRouting

New in Claude Code: /checkup Run /checkup to: 1. Clean up unused skills/MCPs/plugins and save context 2. Dedup your local CLAUDE.md against the checked in CLAUDE.md 3. Break up root CLAUDE.md into nested CLAUDE.md's + skills 4. Turn off slow hooks 5. Update your Claude Code to the latest version 6. Enable auto mode by default 7. Pre-approve frequently denied read-only commands .. And a few other goodies. /checkup confirms with you before making any changes. Enjoy!