# fable Published articles for fable. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## Building a multiplayer game world with autonomous Fable-class models DevFeed: [Building a multiplayer game world with autonomous Fable-class models]() Original publisher: [Read original article]() Author: Simon Willison Published: 2026-10-29T17:00:26Z Content type: opinion Language: en Sources: [Simon Willison's Weblog]() Topics: [multiplayer](), [Continuous Deployment (CD)](), [CI/CD](), [Frontier AI](), [generative-ai-tools]() Tags: [agent](), [agentic](), [agentic-engineering](), [agentic-engineering-70](), [agents](), [ai](), [ai-2-303](), [alert](), [autonomous](), [bug](), [bugs](), [change](), [changes](), [chat](), [claude](), [claude-323](), [claude-mythos-fable](), [claude-mythos-fable-43](), [code](), [coding-agents](), [coding-agents-264](), [coherent](), [continuous](), [fable](), [game](), [game-design](), [game-design-30](), [generative-ai](), [generative-ai-2-043](), [llms](), [llms-2-008](), [models](), [multiplayer](), [world]() ### AI overview Spawn is a multiplayer game where players prompt its agent, Savi, to change the world; the excerpt describes a development process built around Fable-powered agents that edit live code and ship hourly. About 60% of work reportedly begins with production errors, Savi-filed bug reports, unusual logs, or feature requests. Agents triage reports, often deliver fixes to affected users within an hour, and can roll back a broken change in 30 seconds. Alerts wake incident-owning agents, and the team deliberately tests recovery paths. The account says Fable 5 and 5.1 enabled this autonomy, while Opus 5.5 fell short and GPT-6 Astra disabled product features while applying security fixes. The source frames this as an experiment suited to game development, while expressing doubt about applying it to higher-stakes software. ### Source excerpt How I Built the Most AI-Pilled Company in the World Don't be put off by the clickbait headline, this piece by Jacob Sansbury describing how they are building Spawn is a fascinating exploration of the bleeding edge of what's possible with agentic engineering against frontier models. Spawn is a massively multiplayer gaming environment where players can use prompts (to the in-game agent, Savi) to modify the game world. I fired it up on my phone for a few minutes and was able to prompt a castle into existence, then a dragon, then ride the dragon to fly around that castle. By Jacob's telling, Spawn is now being developed autonomously by a swarm of Fable-powered agents costing over $3 million in tokens per month ("we spend 30x payroll on AI tokens"). Our agents change the live code directly and ship to production hourly, so they read every error, bug report and metric themselves. About 60% of our work starts from a production error, a Savi bug report (Savi is the AI you build games with in Spawn), something odd in our logs, or a feature request. Savi files bugs herself. When a user hits one, or Savi hits one in the product, she files it. It goes straight to an agent that decides whether it's a real bug to fix, something we won't fix, or intentional design that needs a decision. Usually it becomes a change, delivered to that user within an hour, with a message in the same chat. This is continuous deployment turned up to 11: A loop this tight only works if it never stops, so we treat an hour without a ship as a SEV-0, our most severe kind of incident. Unshipped changes pile up, and every problem already in production lasts longer. The feedback slows down too. An agent can only keep a few hours of work in its head, so shipping faster than that actually helps it stay coherent. [...] We can roll back a broken change in 30 seconds, exactly where it's broken and nowhere else. When a monitor or a log alert fires, it starts or wakes an agent that owns the incident, and we break th ## JustGoal: With Only a Goal, the AI Baseline Is Above the Expert Kernels DevFeed: [JustGoal: With Only a Goal, the AI Baseline Is Above the Expert Kernels]() Original publisher: [Read original article]() Author: Yaojian Chen, Dazhi Peng, Zhui Zhu, Leshan Li, Haojun Sun, Yiyun Zheng, Qingfeng He, Shangzhan Li, Xianglong Deng, Yifei Shen, Shengye Xiang, Zhen Li, Lin Gan, Yuxuan Li, Xu Han, Zhiyuan Liu Published: 2026-10-28T04:00:00Z Content type: article Language: en Sources: [arXiv: Distributed Computing]() Topics: [LLM optimization](), [AI Agent](), [LLM evaluation / benchmarking]() Tags: [agent](), [ai](), [cs-dc](), [domain](), [efficiency](), [evaluation](), [fable](), [fable-5-1](), [foundation-model](), [high-performance](), [inference](), [infrastructure](), [kernel](), [kernels](), [knowledge](), [language-model](), [libraries](), [operator](), [operators](), [optimization](), [performance]() ### AI overview JustGoal tests whether a general-purpose agent given only an optimization goal and constraints can rival GPU kernel tools with specialized knowledge and workflows. On NVIDIA B200 GPUs, it beats open-source libraries on 8 of 10 operators, with per-operator geometric mean speedups of 1.14x to 2.34x, and exceeds the fastest publicly released specialized-agent kernel on 5 of 9 comparable operators. In a same-model, single-shape comparison using Fable 5.1, the KDA-0.5 harness reaches 0.95x of JustGoal's performance; single-mechanism ablations show no consistent benefit from domain knowledge or predefined workflows. The authors argue specialized machinery should demonstrate gains over a goal-only baseline with the same model and scope, and be reevaluated as foundation models improve. ### Source excerpt arXiv:2610.36434v1 Announce Type: new Abstract: A small number of GPU kernels often determines the efficiency of large language model training and inference. High-performance kernel development typically relies on expert knowledge through manual implementation, collaboration with general-purpose models, or specialized agents equipped with domain knowledge, evaluation methods, and search workflows. We investigate how much value this knowledge provides as foundation models become stronger. We introduce JustGoal, a goal-only baseline in which a general-purpose agent receives only the optimization objective and constraints, without kernel-specific infrastructure, knowledge bases, or predefined workflows, and builds its own evaluation loop. On NVIDIA B200 GPUs, JustGoal outperforms open-source libraries on 8 of 10 operators, achieving per-operator geometric mean speedups of 1.14x to 2.34x. It also surpasses the fastest publicly released specialized-agent kernel on 5 of 9 operators with available comparisons. In a controlled same-model comparison using Fable 5.1 on single shapes, the KDA-0.5 harness reaches 0.95x of JustGoal's performance, while single-mechanism ablations show no consistent benefit from domain knowledge or predefined workflows. These results indicate that specialized machinery should be justified by clear gains over a same-model, same-scope goal-only baseline and reevaluated whenever the foundation model changes. ## Claude Opus 5.5 vs. Fable 5.1: Coding accuracy, benchmark results, and cost DevFeed: [Claude Opus 5.5 vs. Fable 5.1: Coding accuracy, benchmark results, and cost]() Original publisher: [Read original article]() Author: Jessica Wachtel Published: 2026-09-29T15:00:00Z Content type: comparison Language: en Sources: [The New Stack]() Topics: [Claude](), [Frontier AI](), [Test-driven development](), [Concurrency](), [Deadlock](), [data migrations](), [Deep neural networks]() Tags: [ai-agents](), [ai-models](), [anthropic](), [claude](), [claude-fable-5](), [claude-fable-5-1](), [claude-opus](), [claude-opus-5](), [claude-opus-5-5](), [concurrency](), [concurrency-bugs](), [cursorbench](), [fable](), [fable-5-1](), [opus-5-5](), [python](), [race-conditions](), [software-testing]() ### AI overview The author compares Claude Opus 5.5 and Claude Fable 5.1 on coding tasks using identical API prompts, adaptive thinking, and maximum effort. Across repeated tests, the evaluation tracks hidden-test accuracy, token use, cost, and run time across bug fixing, dependency resolution, and concurrency bugs. The excerpt describes the test setup but does not include the results. ### Source excerpt Anthropic released Claude Opus 5.5 on September 22, saying it performs at the level of Claude Fable 5.1 on most The post Claude Opus 5.5 vs. Fable 5.1: One overthinks, the other cuts corners appeared first on The New Stack. ## Claude models: Fable vs. Opus vs. Sonnet vs. Haiku DevFeed: [Claude models: Fable vs. Opus vs. Sonnet vs. Haiku]() Original publisher: [Read original article]() Author: Harry Guinness Published: 2026-09-24T04:00:00Z Content type: comparison Language: en Sources: [Zapier]() Topics: [Claude](), [Frontier AI](), [llm-reasoning](), [AI-assisted coding]() Tags: [2026](), [all](), [also](), [anthropic](), [but](), [claude](), [doesn-t](), [fable](), [family](), [general](), [haiku](), [i](), [june](), [just](), [model](), [models](), [mythos](), [october](), [opus-5](), [productivity](), [vs]() ### AI overview In the author's five prompt tests, Opus gave the most consistent balance of accuracy and cost: it handled ambiguity, proved an impossible scheduling puzzle unsolvable, followed the rewrite constraints, prioritized document contradictions, and built a functional Kanban board. Sonnet was a capable conversational default, with strong document analysis and reasoning at lower cost, though it missed a word count and produced a less polished board. Haiku was fastest and cheapest, and worked well for simple categorization, but struggled with reasoning and silently altered a quote. Fable performed about as well as Opus, with the strongest board design, but costs more, can overthink simple tasks, and may fall back to Opus on requests that trigger safety guardrails. The author recommends Opus for a paid subscriber's general use and Sonnet for free users, token conservation, or work that does not require advanced coding. Findings reflect the author's tests and experience, not comprehensive coverage. ### Source excerpt Anthropic's Claude family has four primary models: Haiku, Sonnet, Opus, and Fable. The version numbers are inconsistent between them; as I write this, Opus 5.5 just dropped, Fable 5.1 came out on September 1, 2026, Sonnet 5 was released on June 30, 2026, and Haiku is still on 4.5 and launched in October 2025--though a new model is due out soon. There's also Mythos, but the general public doesn't have access to that. All four Claude models are incredibly capable. Even the smallest model, Haiku, i ## Bun Rewrites 535K Lines of Zig into Rust in Four Months, Eliminates Numerous Memory Leaks DevFeed: [Bun Rewrites 535K Lines of Zig into Rust in Four Months, Eliminates Numerous Memory Leaks]() Original publisher: [Read original article]() Author: Bruno Couriol Published: 2026-09-20T19:51:00Z Content type: news Language: en Sources: [InfoQ]() Topics: [Bun](), [Rust](), [Memory Safety](), [Vulnerabilities](), [rust-ai](), [TypeScript](), [anthropic](), [JavaScript](), [Fable]() Tags: [agentic](), [ai](), [ai-assisted-coding](), [bun](), [bun-ai-rewrite-zig-rust-4-months](), [bundlers](), [claude](), [code-reviews](), [compiler](), [development](), [fable](), [javascript](), [memory-safety](), [news](), [port](), [porting](), [rust](), [system-programming](), [typescript](), [vulnerabilities](), [web-development]() ### AI overview Bun creator Jarred Sumner says the JavaScript and TypeScript runtime, bundler, and package manager was rewritten from Zig to Rust in four months with AI assistance. The rewrite aimed to address recurring memory-safety vulnerabilities, using Rust's borrow checker and validation against Bun's existing TypeScript test suite. ### Source excerpt Bun creator Jarred Sumner recently announced that Bun, the JavaScript/TypeScript runtime, bundler and package manager has been rewritten from Zig to Rust. The rewrite seeks to eliminate recurrent memory safety vulnerabilities via Rust's borrow checker. The AI-assisted rewrite was released after 4 months of work instead of the estimated one year. By Bruno Couriol ## AI's best coding agent fails 60% of the time -- and the data backs it up DevFeed: [AI's best coding agent fails 60% of the time -- and the data backs it up]() Original publisher: [Read original article]() Author: Amanda Caswell Published: 2026-09-14T22:22:27Z Content type: news Language: en Sources: [The New Stack]() Topics: [Benchmark](), [AI-assisted coding](), [Fable](), [Claude Code](), [codex](), [Command-line interface](), [anthropic]() Tags: [ai](), [ai-engineering](), [ai-models](), [benchmark](), [benchmarks](), [claude-code](), [cli](), [codex](), [coding](), [fable](), [performance](), [software-testing]() ### AI overview Real-SWE evaluates coding agents on private company codebases and reports substantially lower success rates than public-repository benchmarks. Claude Fable 5.1, running through Claude Code, led the comparison with a 38.8% score, while the tested systems often failed most attempts. ### Source excerpt Claude Fable 5.1 just won a new coding benchmark despite failing more than six out of 10 times. Its 38.8% The post AI's best coding agent fails 60% of the time -- and the data backs it up appeared first on The New Stack. ## How Fable and Astra Could Change Software Engineering Work DevFeed: [How Fable and Astra Could Change Software Engineering Work]() Original publisher: [Read original article]() Author: Hemant Pandey Published: 2026-09-09T14:09:16Z Content type: opinion Language: en Sources: [The Hustling Engineer]() Topics: [Artificial Intelligence](), [Programming](), [Fable](), [computer-use](), [Code](), [anthropic](), [OpenAI]() Tags: [ai](), [anthropic](), [code](), [coding](), [computer-use](), [engineering](), [fable](), [openai](), [terminal](), [vs-code]() ### AI overview This opinion article examines what Fable and Astra could change for software engineers. It contrasts Fable's long-running, multi-step coding and research work with Astra's ability to operate software and complete tasks across tools used by engineers. ### Source excerpt What actually changes for software engineers after Fable and Astra? ## Testing Astra 6 v Fable 5.1 on a Gradle docs bug DevFeed: [Testing Astra 6 v Fable 5.1 on a Gradle docs bug]() Original publisher: [Read original article]() Author: Laura Kassovic Published: 2026-09-09T04:00:00Z Content type: opinion Language: en Sources: [Gradle]() Topics: [Gradle](), [Software Testing](), [AI-assisted coding](), [Documentation](), [gpt-6-astra](), [codex](), [Fable]() Tags: [agentic](), [agents](), [ai](), [ai-coding](), [api](), [astra](), [benchmark](), [bug](), [cli](), [coding](), [cost](), [docs](), [documentation](), [dsl](), [fable](), [fable-5-1](), [git](), [github](), [gradle](), [testing]() ### AI overview The Gradle docs bug came from Dokka deriving Java interop links from the JDK 21 build toolchain, even though Gradle targets Java 17; Kotlin DSL references therefore pointed to Java 21 Javadoc instead of Java 17. The post compares claude-fable-5-1 via claude-code with gpt-6-astra via codex on the same issue and constraints. Both fixes passed an independent judge, but the author evaluates cost, speed, and maintainability as separate dimensions. The evidence is limited to one run per model, one judge, and one bug, so it does not support a broad benchmark conclusion. ### Source excerpt On the afternoon of September 7th, claude-fable-5-1 was handed a git repository, a GitHub issue, and fifty turns to fix it. Just over three hours later, gpt-6-astra finished the same assignment in another worktree. Same issue. Same constraints. Different CLI wrapped around each model: claude-code for Fable, codex for Astra. Both Agents fixed the bug. Both results were graded as successes by an independent judge. But these results were quite different. You would be forgiven for wondering why an engineering blog is running a two-model bakeoff on a single documentation bug, in a year when everybody and their intern has published an AI coding benchmark. Here's the honest answer: we wanted to know which one actually did the better job, as a side quest of the Agentic Gradle project. But "better" was never going to mean pass-or-fail. Both of these agents passed; if that were the whole story, this post would be four sentences long. What actually separates a fix worth merging from a fix worth sending back is cost, speed, and whether the result is something we would want to maintain, and those three do not all point the same direction, as you are about to see. So: one bug, two agents, and a very literal stopwatch. Here is what this looks like when you actually read the agents' transcripts. Real talk up front: this is n=1. One run per model, one judge, one bug. That's a thin base for anything you'd call a leaderboard, and I'll say so again at the end, but it didn't stop me from landing on an opinion by the time I'd finished reading both transcripts. The bug, briefly Issue #34751, filed August 21st by our very own cobexer, is the kind of bug that is very easy to describe and mildly annoying to fix. In the Kotlin DSL API docs, a type like Attribute linked to the Java 21 Javadoc. It should have linked to Java 17, because Gradle targets Java 17. The reason is almost embarrassingly mundane once you see it. Gradle is built with a JDK 21 toolchain but targets JDK 17, which i ## Claude Fable 5.1: Benchmark results and reasoning-effort experiments DevFeed: [Claude Fable 5.1: Benchmark results and reasoning-effort experiments]() Original publisher: [Read original article]() Author: Simon Willison Published: 2026-09-01T23:57:28Z Content type: article Language: en Sources: [Simon Willison]() Topics: [Fable](), [Benchmark](), [Large Language Model](), [Programming](), [SVG]() Tags: [ai](), [ai-2-235](), [anthropic](), [anthropic-336](), [benchmark](), [benchmarks](), [claude](), [claude-310](), [coding](), [fable](), [generative-ai](), [generative-ai-1-981](), [llm](), [llm-reasoning](), [llm-reasoning-103](), [llm-release](), [llm-release-231](), [llms](), [llms-1-947](), [models](), [pelican-riding-a-bicycle](), [pelican-riding-a-bicycle-142](), [reasoning](), [svg]() ### AI overview The article examines Claude Fable 5.1 through benchmark results and a pelican SVG-generation experiment across its five reasoning-effort levels. It reports that low and medium settings appeared to skip reasoning for this prompt, while higher effort used more tokens, time, and cost. ### Source excerpt Today is Claude Fable (and Mythos) 5.1 day. Anthropic say that Fable 5.1 "sets a new standard for coding, knowledge work, and long-running problem-solving tasks". Their announcement spends a notable amount of time on scientific research, boasting of a 52.6% score on the brand new Terminal-Bench-Science 0.1 benchmark (first announced on August 27th), up from 24.7% for Fable 5, 29.0% for Opus 5 and 22.4% for GPT-5.6 Sol. Other benchmarks show slightly improved scores, but none as impressive as the Science one. But how well can it pelican? Back in July I wrote about how I was losing faith in the pelican benchmark - its connection to how good the models were at other tasks didn't seem to hold as strongly as it did back in 2025. The most interesting insights I get from it now are comparisons within model families, and particularly comparisons for the same prompt at different reasoning effort levels. Fable 5.1 has five reasoning levels: low, medium, high, xhigh, max - and no option to turn off reasoning entirely. I fixed an issue in llm-anthropic which caused reasoning traces not to be correctly recorded, then ran some prompts. Here's the full set of pelicans for all of the reasoning levels, each with the full reasoning transcript. I'll replicate them here: Low and medium, both without reasoning? Next, a bit of a mystery. This is what I got for effort low: The transcript doesn't show any summarized reasoning tokens, and the output token count is 1,998. With Claude that output token count includes reasoning tokens. It took 23.8 seconds and cost 10.017 cents. I bumped that up to medium and got this: Weirdly, that one also shows no reasoning text and used 1,977 output tokens - 21 tokens less than low. It took 23 seconds and cost 9.912 cents. So for this particular prompt ("Generate an SVG of a pelican riding a bicycle") Fable 5.1 appeared to skip reasoning entirely at both low and medium settings. High Here's high - 29.6 seconds, 2,612 output tokens, 13.087 cents: This one d ## Fable 5.1 review: Coding tests and code review results DevFeed: [Fable 5.1 review: Coding tests and code review results]() Original publisher: [Read original article]() Author: Juan Pablo Flores Published: 2026-09-01T00:00:00Z Content type: news Language: en Sources: [CodeRabbit]() Topics: [AI-assisted coding](), [Code review](), [ai code review](), [agentic-coding](), [Test automation](), [reproducibility]() Tags: [ai](), [arena](), [automated](), [autonomous](), [code](), [code-review](), [coding](), [coding-agents](), [coding-tasks](), [english](), [fable](), [fable-5](), [fable-5-1](), [latency](), [models](), [results](), [review](), [switching](), [test](), [tests](), [trade]() ### AI overview Fable 5.1 is best suited to complex coding changes and selective code review, where broader context and planning matter more than speed. In CodeRabbit's 45-task review evaluation, it found one fewer of 105 known issues than Fable 5, while producing 87 fewer final comments and improving precision by 4.5 percentage points. Precision was still 37.3%, and average review time rose to 18 minutes 38 seconds from 12 minutes 32 seconds; the comparison used different review-system versions. Low reasoning outperformed High overall: it found more known issues and took nearly three minutes less per task, with the same call count. In coding tests, Fable 5.1 checked its environment and planned before editing, handled simple tasks quickly, and stayed close to underspecified requests. A complex Snake arena test showed it could coordinate five strategies and shared browser and command-line behavior, though this is a single example. CodeRabbit recommends testing representative repository tasks and measuring useful comments, issue coverage, latency, and output volume before changing defaults. ### Source excerpt Read CodeRabbit's Fable 5.1 review, including coding tasks, code review results, latency, and the trade-offs to test before switching models. ## Announcing some improvements to \`renovate-pretty-log-tui\` DevFeed: [Announcing some improvements to \`renovate-pretty-log-tui\`]() Original publisher: [Read original article]() Author: Jamie Tanna Published: 2026-08-17T20:47:36Z Content type: release Language: en Sources: [Jamie Tanna]() Topics: [renovate](), [Text-based user interface](), [releases](), [debugging](), [Claude](), [User Experience](), [Fable](), [HTML](), [browser](), [GitHub](), [file]() Tags: [browser](), [claude](), [debug](), [fable](), [github](), [html](), [improvements](), [release](), [release-notes](), [renovate](), [ux]() ### AI overview An update to renovate-pretty-log-tui and related utilities improves the experience of viewing Renovate debug logs. Version 0.6.0 adds a summary view, clearer error presentation, better support for smaller displays, and an option to export logs as a self-contained HTML page. ### Source excerpt As I've blogged about before, I've written a tool for improving the experience of working with Renovate's debug logs, providing a more targeted experience for viewing the logs. While there are a few options out there at the moment, including part of our products at Mend, it's also nice having an agnostic tool I can use across different products, or where a customer or a user has provided logs to us. Last week, I finally sat down to listing out some of the key improvements I'd like to have completed on renovate-pretty-log-tui. As I'm not that familiar with Bubble Tea, which it's based on, I decided to set Claude Opus and Claude Fable on the work, as well as seeing if they could see any additional improvements. I've just shipped v0.6.0 of my renovate-pretty-log-tui and renovate-pretty-log utilities, including a whole host of changes - check out the release notes for more. Some of the key changes can be seen in this Asciicast: As you can see, there's now a summary view, and an improved UX for viewing the logs. I'm still working on how this works on smaller screens (i.e. in a Tmux pane) but it works nicely when it's rendered on a half-screen window size. In this, we also surface errors more clearly, i.e. Although I'd mostly completed this last week, I needed to put the last touches to the commit messages. With the GitHub outage today, I decided to finish getting these changes complete, as they've very useful for working with debug logs. I'd also decided to add renovate-pretty-log -html option, which allows converting a debug log into a single-page, self-contained HTML file that allows reviewing a very similar view of the local tools, but in the browser. ## One-shotting a Raccoon Heist game using Claude Fable 5 DevFeed: [One-shotting a Raccoon Heist game using Claude Fable 5]() Original publisher: [Read original article]() Author: Simon Willison Published: 2026-08-05T19:42:38Z Content type: tutorial Language: en Sources: [Simon Willison]() Topics: [Claude](), [Claude Code](), [GitHub Pages](), [GitHub](), [Code]() Tags: [ai](), [ai-2-235](), [anthropic](), [anthropic-336](), [browser](), [claude](), [claude-310](), [claude-code](), [claude-mythos-fable](), [claude-mythos-fable-41](), [code](), [coding-agents](), [coding-agents-248](), [fable](), [game-design](), [game-design-22](), [generative-ai](), [generative-ai-1-981](), [github](), [github-pages](), [iphone](), [llms](), [llms-1-947](), [prompt-engineering](), [prompt-engineering-197](), [text-to-image](), [text-to-image-47](), [vibe-coding](), [vibe-coding-97]() ### AI overview The article describes an experiment in which Claude Fable 5, running through Claude Code for web, built a Raccoon Heist game from an older GPT-3 game concept and screenshots. It also explains using GitHub Pages to preview work from Claude Code for web and notes that published content from a private repository may be viewable by anyone who guesses the repository name. ### Source excerpt Back in 2022 I tweeted screenshots of a game concept generated by GPT-3 and some concept "art" created using DALL-E. Today, on the fourth anniversary of that tweet, I decided to see if Claude Fable 5 (running in Claude Code for web) could build the entire game from the content of that tweet. It did a pretty good job of it! You can play the game here. Here's the GitHub repo, and a short video demo: Your browser does not support HTML5 video. How I built this This is the August 5th, 2022 tweet: My GPT-3 prompt back then was: Write a detailed product description of a computer game where a team of raccoons go on heists. This was in the text completion days, where you asked GPT-3 to complete the rest of the text. It completed: In "Raccoon Heist", you and your team of thieving raccoons are tasked with pulling off a series of daring heists. From robbing banks to stealing priceless art, no job is too big or too small for your furry crew. You'll need to use your wits and your skills to avoid the police and make a clean getaway with the loot. With exciting gameplay and a charming cast of characters, "Raccoon Heist" is the perfect game for anyone looking for a light-hearted caper. My prompt to DALL-E was short: Screenshot from a video game where a team of raccoons go on a heist Today's experiment: can I dump those screenshots into Fable 5 with a prompt telling it to write a game, then leave it to its own devices and get a working game at the end? Setting Claude Code for web up to use GitHub Pages A frustrating thing about Claude Code for web is that it can be hard to test what it's working on while it's still working. I've been using GitHub Pages to work around that limitation, and found it to work really well. Here's my process: Create a new repository for the project at https://github.com/new - this can be public or private, the trick works equally well for both. Start a Claude Code for web session, in the Claude iPhone or Desktop apps or in the browser at https://claude.ai/c ## How to code with Kimi K3 better than Fable and no restrictions DevFeed: [How to code with Kimi K3 better than Fable and no restrictions]() Original publisher: [Read original article]() Author: Pieter Levels Published: 2026-07-17T00:00:00Z Content type: tutorial Language: en Sources: [@levelsio (Pieter Levels)]() Topics: [codex](), [Claude Code](), [Jailbreak]() Tags: [account](), [ai](), [api-key](), [build](), [bypass](), [claude](), [claude-code](), [code](), [codex](), [fable](), [how-to](), [install](), [k3](), [kimi](), [kimi-k3](), [opencode](), [permissions](), [run](), [settings](), [tech]() ### AI overview The post gives a setup sequence for using Kimi Code with OpenCode: ask Codex or Claude Code to install OpenCode, create a Kimi account, pay for the stated $19/month membership, obtain an API key from kimi.com/code/console, then run OpenCode and use `/connect` to connect to Kimi Code with that key. The author recommends BUILD mode, selected with Shift+Tab, and says bypass permissions can be enabled in `/settings`. These are the post's instructions and preferences; it provides no comparison evidence for the title's claim about Fable. ### Source excerpt How to code with Kimi K3 (better than Fable and no restrictions) today: Ask Codex or Claude Code to install OpenCode (yes it will 😂!) Go to kimi.com, make an a... ## Bun's migration from Zig to Rust reduced the estimated timeline from 1-2 years to 11 days DevFeed: [Bun's migration from Zig to Rust reduced the estimated timeline from 1-2 years to 11 days]() Original publisher: [Read original article]() Author: Ivan Klaric Published: 2026-07-16T16:50:20Z Content type: opinion Language: en Sources: [The Pragmatic Engineer]() Topics: [Bun](), [migration](), [Rust](), [Zig](), [Fable](), [Memory Leaks]() Tags: [bun](), [fable](), [memory-leaks](), [migration](), [rust](), [zig]() ### AI overview The article examines Bun's rewrite from Zig to Rust, focusing on memory-safety problems and the migration approach. It reports that the migration timeline was shortened from an estimated 1-2 years to 11 days, while emphasizing that thorough testing was required. ### Source excerpt To a sceptic, spending $165K to migrate Bun from Zig to Rust sounds very expensive. But to a realist, shortening a 1-2 year migration down to 11 days opens amazing new opportunities for devs. However, a thoroughly-tested project is required to pull it off. ## Athena's free vulnerability intelligence model and the economics of security clearinghouses DevFeed: [Athena's free vulnerability intelligence model and the economics of security clearinghouses]() Original publisher: [Read original article]() Published: 2026-07-15T00:00:00Z Content type: opinion Language: en Sources: [Chainguard: Unchained]() Topics: [Security](), [Open Source](), [Critical Infrastructure](), [Network]() Tags: [akrites](), [chainguard-athena](), [critical-infrastructure](), [fable](), [infrastructure](), [money](), [mythos](), [network](), [open-source](), [security](), [software-supply-chain-security](), [trust](), [vulnerability-clearninghouse]() ### AI overview The article explains why Athena provides pre-disclosure vulnerability findings to infrastructure and network partners for free. It argues that allowing partners to commercialize services built on advance security information can align incentives, trust, and broader defensive value, while distinguishing government programs from commercial clearinghouses. ### Source excerpt Why give away the most valuable data in security? Learn how Athena's business model aligns trust, incentives, and open source defense. ## Which Doc Format is Best for AI Specifications? DevFeed: [Which Doc Format is Best for AI Specifications?]() Original publisher: [Read original article]() Author: Peter Lawrey (noreply@blogger.com) Published: 2026-07-14T14:00:28Z Content type: article Language: en Sources: [Vanilla Java]() Topics: [Specifications](), [Markdown](), [Fable](), [HTML](), [Documentation]() Tags: [ai](), [fable](), [html](), [info](), [markdown](), [readability](), [specifications]() ### AI overview The article compares Markdown, AsciiDoc, and HTML for AI specifications using document-size, markup-overhead, readability, and editing-workflow observations. It recommends Markdown for AI working documents, AsciiDoc for curated human-reviewed specifications, and HTML primarily as a publishing target. ### Source excerpt I have a Spec Driven Project with 674 documents incl 181 specs, 429 tasks, and 40 project docs. About 1/3 of these are an acceptance sub-project to show the root project meets fit-for-purpose requirements. I used Fable to convert these to and from AsciiDoc, Markdown, and HTML. From this, I tried to draw some conclusions about which format is better than the others for AI specifications. TL;DR: Markdown for AI working documents, AsciiDoc for curated human-reviewed specs, HTML only as a publishing target. Objective Metrics Corpus Tokens (Approx) Markdown: 594k, AsciiDoc: 598K, HTML: 661k Worst Single-File Ratio AsciiDoc: 1.023x, HTML: 1.26x (due to tables and code blocks) Markup share of file content Markdown 0.9%, AsciiDoc 1.3%, HTML 9.9% - a 10x overhead for HTML Read/write Assessment In terms of readability, both Markdown and AsciiDoc are good. HTML has far more tags, which add noise. One of the most common tags is some code however in the other formats, this is just `some code` When you have an average of 16 per file, this adds a lot of noise. Usability GitHub renders both Markdown and AsciiDoc with working links and checkboxes; HTML is shown as raw source, so the review loop dies there. The first two are also easier to view while you're editing in IntelliJ. HTML is more difficult to read and edit, and it is not as easy to view the rendered output. If you occasionally need HTML, you don't need HTML files: embed an island with ```{=html} in Markdown (Pandoc raw-attribute syntax) or in AsciiDoc. It renders where HTML is supported and drops out cleanly elsewhere. If you really need to use HTML, you can include a link to a file in that format. Note As these were one-for-one conversions, I didn't use features that only exist in AsciiDoc or HTML, such as. Markdown vs AsciiDoc While you might choose either format for your specs, I have been following this convention to keep the authoring consistent. Markdown used for AI-generated content, not human-written, ## Dodo Digest: Even Anthropic Has To Answer To Governments DevFeed: [Dodo Digest: Even Anthropic Has To Answer To Governments]() Original publisher: [Read original article]() Author: Rishabh Goel Published: 2026-07-02T00:00:00Z Content type: opinion Language: en Sources: [Dodo Payments Blog]() Topics: [anthropic](), [Claude]() Tags: [accessible](), [adaptive](), [agents](), [ai](), [ai-companies](), [anthropic](), [building](), [challenge](), [dodo](), [fable](), [government](), [handling](), [merchant-of-record](), [models](), [mythos](), [newsletter](), [payments](), [processor](), [product](), [regulatory](), [startups]() ### AI overview The newsletter uses the U.S. government's clearance of Anthropic's Mythos and Fable models to highlight that global product sales bring regulatory and operational demands alongside building the product. It presents Merchant of Record infrastructure as a way for startups to handle payments and compliance without building that operation themselves. Dodo Payments v1.105.0 adds Bring Your Own Processor, allowing businesses to retain existing processor relationships while using Dodo's billing, subscriptions, tax handling, and Merchant of Record capabilities. It also adds decimal-aware handling for JPY, KRW, VND, CLP, XAF, and XOF across payments, invoices, payouts, emails, analytics, and currency conversions, aiming to avoid rounding issues. ### Source excerpt Anthropic's Mythos and Fable models are back after the U.S. government lifted restrictions, a reminder that even the largest AI companies operate within regulatory boundaries. Building a product is one challenge; selling it globally is another, which is why Merchant of Record models matter for startups. Plus, we shipped Dodo Payments v1.105.0 with Bring Your Own Processor (BYOP) and decimal-aware currency handling. ## Claude Fable 5 access restored on AI Gateway DevFeed: [Claude Fable 5 access restored on AI Gateway]() Original publisher: [Read original article]() Author: Jerilyn Zheng Published: 2026-07-01T07:01:00Z Content type: release Language: en Sources: [Vercel News]() Topics: [Fable](), [Claude](), [API](), [Artificial Intelligence](), [debugging]() Tags: [anthropic-claude](), [api](), [coding](), [debugging](), [docs](), [fable](), [government](), [model](), [models](), [real-time](), [retention](), [safety]() ### AI overview Access to the Claude Fable 5 Mythos-class model has been restored on AI Gateway after the US Government lifted export controls. The release updates its safety classifiers and documents model fallbacks, API usage, and 30-day retention of prompts and completions. ### Source excerpt Access to Claude Fable 5, the Mythos-class model, has now been restored on AI Gateway following the US Government's decision to lift the export controls. Fable 5 is the same model that was available between June 9 and June 12. What has changed is the safety classifiers, which are now updated and more robust. In the near term, some routine tasks such as coding and debugging may trigger safety classifiers. To ensure requests are still serviced when the safety classifiers are triggered, use model fallbacks. AI Gateway will try each model in models in the stated order if Anthropic refuses the request to Fable 5. To call Fable 5, use model name anthropic/claude-fable-5: Model fallbacks work on every API format: for more information on how to configure these, see the docs. Anthropic does not support Zero Data Retention for the model, because some misuse patterns are only visible across cumulative requests, which real-time filters cannot catch on their own. Prompts and completions are retained for 30 days and are not used to train Claude. Read more in the data retention whitepaper. Read more ## From Harness Lock-In to Portable Context Layer DevFeed: [From Harness Lock-In to Portable Context Layer]() Original publisher: [Read original article]() Author: Paul Iusztin Published: 2026-06-30T08:20:17Z Content type: article Language: en Sources: [Decoding ML]() Topics: [AI Context Management](), [Model Context Protocol](), [Agent Harness](), [vendor lock-in](), [Agent Skills](), [Retrieval Augmented Generation (RAG)]() Tags: [agent](), [ai-assistant](), [architecture](), [bm25](), [business-logic](), [claude-code](), [context-layer](), [conversations](), [fable](), [filesystem](), [google](), [harness](), [interface](), [knowledge-graph](), [mcp-servers](), [skills](), [unified-memory]() ### AI overview The article argues that keeping personal and business context separate from an AI harness reduces switching costs when models, tools, or pricing change. It describes a three-part setup: unified memory, a serving layer, and a disposable harness. The memory combines a filesystem, BM25 keyword search, semantic vector search, and a knowledge graph of typed Person, Organization, Location, Event, and Object entities. MCP servers or skills expose that memory and its business logic to agents. The author cites Scrabble, connected to Obsidian, Readwise, Notion, and Google tools, and a separate personal-assistant project. The excerpt does not provide the rest of the architecture or evidence for the promised five-minute setup. ### Source excerpt Build a unified memory (knowledge graph or an LLM wiki) and serve it over MCP servers or skills, so any agent, open or closed, plugs in within minutes. ## Scaling Context Management: Comparing Fable 5 and GPT-5.5 DevFeed: [Scaling Context Management: Comparing Fable 5 and GPT-5.5]() Original publisher: [Read original article]() Author: Calvin French-Owen Published: 2026-06-13T12:00:00Z Content type: opinion Language: en Sources: [Calvin French-Owen]() Topics: [long-context](), [codex](), [Fable](), [API](), [Cache](), [GPU]() Tags: [api](), [caching](), [codex](), [context](), [fable](), [gpu](), [long-context]() ### AI overview The article compares how Fable 5 and GPT-5.5 scale context management for difficult, long-running tasks. It describes Codex's server-side compaction approach and discusses its implications for clients and long-running threads. ### Source excerpt Comparing two very different approaches to scaling context between Fable 5 and GPT 5.5. ## What Anthropic Released with Claude Fable 5 and Its Cybersecurity Limitations DevFeed: [What Anthropic Released with Claude Fable 5 and Its Cybersecurity Limitations]() Original publisher: [Read original article]() Author: Mike Wilkes Published: 2026-06-13T03:06:00Z Content type: news Language: en Sources: [Aikido Security's Blog]() Topics: [anthropic](), [Security](), [Fable](), [Cybersecurity](), [Vulnerabilities]() Tags: [anthropic](), [business-logic-flaws](), [cybersecurity](), [fable](), [infosec](), [news]() ### AI overview The article examines Anthropic's release of Claude Fable 5, arguing that its routing of cybersecurity queries to another model makes direct benchmarking of its security capabilities ineffective. It also discusses Mythos Preview access and the continuing role of people in software security. ### Source excerpt You never needed Mythos to find your IDORs and business logic flaws. A look at what Anthropic shipped with Fable 5, and why infosec stays a people problem at heart. Category: News ## I revived a 28 year old Quake 2 map with AI into web-based multiplayer DevFeed: [I revived a 28 year old Quake 2 map with AI into web-based multiplayer]() Original publisher: [Read original article]() Author: Pieter Levels Published: 2026-06-11T00:00:00Z Content type: release Language: en Sources: [@levelsio (Pieter Levels)]() Topics: [multiplayer](), [Web]() Tags: [ai](), [fable](), [games](), [map](), [multiplayer](), [quake](), [stuff-i-made](), [tech](), [web](), [web-based]() ### AI overview A 28-year-old custom Quake 2 map by @javilopen was revived as a web-based multiplayer game using Fable in fast mode. The first attempt failed to load textures; repacking some PAK files fixed the problem. The author presents the project as an example of using AI to explore and revive old files from backups. ### Source excerpt I have revived @javilopen's 28 year old custom map he made and made a web-based Quake 2 with Fable on fast mode 🤓 You can now play it here: 👉 q2.pieter.com 👈 I... ## I ported Return to Castle Wolfenstein to the web with multiplayer in an hour using AI DevFeed: [I ported Return to Castle Wolfenstein to the web with multiplayer in an hour using AI]() Original publisher: [Read original article]() Author: Pieter Levels Published: 2026-06-11T00:00:00Z Content type: article Language: en Sources: [@levelsio (Pieter Levels)]() Topics: [multiplayer](), [Web]() Tags: [ai](), [cd](), [deflate](), [fable](), [filesystem](), [game](), [games](), [gpl](), [https](), [i386](), [installer](), [iso](), [multiplayer](), [native](), [networking](), [opengl](), [protocols](), [quake](), [tech](), [the-web](), [web](), [wolfenstein]() ### AI overview The author reports porting Return to Castle Wolfenstein's multiplayer mode to the web in about an hour of work, using Fable. The port started from the GPL iortcw source and adapted Emscripten patches from Wwasm, a single-player-only port; because multiplayer code differed substantially, the author manually grafted the relevant changes and fixed renderer initialization. The engine was compiled to WebAssembly with GL4ES translating OpenGL 1.x calls to WebGL2. Multiplayer networking was restored, and a matching native dedicated server was built from the same source. Game data was extracted from the 2001 CD installer, then downloaded by the browser before startup and stored in IndexedDB for reuse. The excerpt describes a working multiplayer game, including the mp_beach map. ### Source excerpt I have to stop boring all of you with my game ports but I ported another game to web with multiplayer, after Quake 1 yesterday and Quake 2 today, I thought, wh... ## Fable 5 vs Opus 4.8: A Data-Driven Decision Guide for Engineering Teams DevFeed: [Fable 5 vs Opus 4.8: A Data-Driven Decision Guide for Engineering Teams]() Original publisher: [Read original article]() Author: Developers Digest Published: 2026-06-10T00:00:00Z Content type: article Language: en Sources: [Developers Digest]() Topics: [Task](), [data]() Tags: [ai-agents](), [ai-models](), [anthropic](), [code-review](), [data-driven](), [developer-tools](), [engineering](), [fable](), [fable-5](), [guide](), [llms](), [swe-bench](), [teams](), [vs]() ### AI overview A decision guide compares Fable 5 with Opus 4.8 using an 80.3% SWE-Bench Pro score, relative cost, and task profiles to assess when the premium may be worthwhile. ### Source excerpt Fable 5 posts an 80.3% SWE-Bench Pro score and costs 2x Opus 4.8 - here is the task-profile scoring guide that tells you when the premium pays off.