Coding Agents Just Changed Formal Verification Forever | Poteto's Skill, Bend2, Drifting Models

Requirements grilling and polished mockups can still miss what happens when a user retries, a request fails, or two actions overlap.Formal verification can surface those gaps before implementation by checking a model of the proposed behavior against promises you've chosen for the product. In this episode, we build on pstack's /statechart skill by Lauren Tan (a.k.a. Poteto) with a workflow that turns counterexamples into product decisions. We explain the model, promises, and checker, illustrate how retries can create duplicate jobs, and show a real calendar flow where the promise itself needed reconsidering. We also cover where this process is worth the effort and why checking a model doesn't guarantee the app will match it.Skill: https://github.com/bkase/plugins/blob/main/pstack/skills/flow-promises/SKILL.mdSuperlinear is a podcast about emerging practices for building with AI.Hosted by Brandon Kase and Christine Yip.Discord: https://superlinear.fm/discordHomepage: ⁠⁠https://superlinear.fm⁠⁠Twitter: ⁠⁠https://x.com/superlinear_fm⁠⁠LinkedIn: ⁠⁠https://www.linkedin.com/company/superlinearfm⁠⁠Brandon Kase: ⁠⁠https://x.com/bkase_⁠⁠Christine Yip: ⁠⁠https://x.com/christinetyip⁠Music licensed through Soundstripe. Code: O2UT6CCEQNH4KT0D

Continue the conversation from the episode on Discord: https://superlinear.fm/discord  The create-formal-verification-skill by Poteto, aka Lauren Tan, can find a valuable property in your codebase, choose a verification tool, and generate a skill future coding agents can use to rerun and maintain the check. Formal verification is becoming a practical part of building with agents.But a passing proof only covers the property and assumptions being checked. If it describes a separate model that has drifted from your application, your shipped code can behave differently. We discuss how to start with Poteto's skill, what Anthropic's recent verification research shows, and how Victor Taelin's Bend2 connects executable code and proofs. We also examine state-machine checking and an experimental approach to bringing verification into the application itself. Superlinear is a podcast about emerging practices for building with AI.Hosted by Brandon Kase and Christine Yip.Discord: https://superlinear.fm/discordHomepage: ⁠⁠https://superlinear.fm⁠⁠Twitter: ⁠⁠https://x.com/superlinear_fm⁠⁠LinkedIn: ⁠⁠https://www.linkedin.com/company/superlinearfm⁠⁠Brandon Kase: ⁠⁠https://x.com/bkase_⁠⁠Christine Yip: ⁠⁠https://x.com/christinetyip⁠Music licensed through Soundstripe. Code: O2UT6CCEQNH4KT0D

An agent that handles new GitHub issues needs to keep working through crashes, restarts, and changes in who's using it. Pi Durable gives you a foundation for building that kind of persistent workflow.The key is separating agent state from the process running it. Cloudflare Durable Objects can give that state a shared cloud home, with execution triggered by requests and configured wakeups.In this episode, we explain how the two fit together, use an interrupted refund to examine recovery, and look at shared chat, runtime reloads, and a phone-local agent. We also discuss what stored task state could reveal without advancing the agent.Superlinear is a podcast about emerging practices for building with AI.Hosted by Brandon Kase and Christine Yip.Discord: https://superlinear.fm/discordHomepage: ⁠⁠https://superlinear.fm⁠⁠Twitter: ⁠⁠https://x.com/superlinear_fm⁠⁠LinkedIn: ⁠⁠https://www.linkedin.com/company/superlinearfm⁠⁠Brandon Kase: ⁠⁠https://x.com/bkase_⁠⁠Christine Yip: ⁠⁠https://x.com/christinetyip⁠Music licensed through Soundstripe. Code: O2UT6CCEQNH4KT0D

Your coding agent can spend more time waiting for tests and builds than writing code. As inference gets faster, tool calls can become a larger share of the wait.Speeding up that feedback loop means examining which checks run, how they run, and where they run. Crabbox, from Peter Steinberger and the OpenClaw team, keeps agents local while expensive work runs on other machines.In this episode, we share our experience offloading Electron tests with Crabbox, discuss the tradeoffs of selective testing and parallel verification, and look at faster tools such as Oxlint and a custom Swift linter. We also discuss which optimizations are worth the effort.Superlinear is a podcast about emerging practices for building with AI.Hosted by Brandon Kase and Christine Yip.Discord: https://superlinear.fm/discordHomepage: ⁠⁠https://superlinear.fm⁠⁠Twitter: ⁠⁠https://x.com/superlinear_fm⁠⁠LinkedIn: ⁠⁠https://www.linkedin.com/company/superlinearfm⁠⁠Brandon Kase: ⁠⁠https://x.com/bkase_⁠⁠Christine Yip: ⁠⁠https://x.com/christinetyip⁠Music licensed through Soundstripe. Code: O2UT6CCEQNH4KT0D

How do you build an agent like Dots, Muse, Instinct, or Grok Bot? Start high up the stack, and move down only when you need more control.YC F26-startup Agent 37 founder, Vishnu Krishnaprasad, explains when to use Hermes, OpenClaw, or Pi, why each user needs a separate sandbox, and how to turn a proven manual workflow into skills. He also demos a Dots-like agent built on Hermes.Superlinear is a podcast about emerging practices for building with AI.Hosted by Brandon Kase and Christine Yip.Discord: https://superlinear.fm/discordHomepage: ⁠⁠https://superlinear.fm⁠⁠Twitter: ⁠⁠https://x.com/superlinear_fm⁠⁠LinkedIn: ⁠⁠https://www.linkedin.com/company/superlinearfm⁠⁠Brandon Kase: ⁠⁠https://x.com/bkase_⁠⁠Christine Yip: ⁠⁠https://x.com/christinetyip⁠Music licensed through Soundstripe. Code: O2UT6CCEQNH4KT0D

Bend2 puts code and proofs in the same language, so coding agents can write both. Tests check examples. A checked proof can establish that a specific rule holds for every allowed input.This is formal verification, an additional verification layer alongside types and tests. You still need to review whether the rule captures the behavior you actually want.In this episode, we show how to choose rules worth proving, get the agent to finish the proofs, and decide where the extra effort pays off. We uncover false proposed rules and integrate a Bend core with checked editing rules into an existing TypeScript app. You can start with one important part of your own project.Superlinear is a podcast about emerging practices for building with AI.Hosted by Brandon Kase and Christine Yip.Discord: https://superlinear.fm/discordHomepage: ⁠⁠https://superlinear.fm⁠⁠Twitter: ⁠⁠https://x.com/superlinear_fm⁠⁠LinkedIn: ⁠⁠https://www.linkedin.com/company/superlinearfm⁠⁠Brandon Kase: ⁠⁠https://x.com/bkase_⁠⁠Christine Yip: ⁠⁠https://x.com/christinetyip⁠Music licensed through Soundstripe. Code: O2UT6CCEQNH4KT0D

Your coding agent does not need a frontier LLM call for every judgment. Deciding which tests to run, which skill to load, or whether a prompt needs more context can be a bounded question.Jev and other decision models return typed choices, scores, or yes/no probabilities quickly enough to sit inside repeated workflows. A valid answer can still be wrong, so placement and fallback rules matter.We show a live prompt checker and a small, six-commit test-selection experiment, then explore agent routing, local Laya, and how to choose between code, a decision model, a generative model, or a human for each step. Superlinear is a podcast about emerging practices for building with AI.Hosted by Brandon Kase and Christine Yip.Discord: https://superlinear.fm/discordHomepage: ⁠⁠https://superlinear.fm⁠⁠Twitter: ⁠⁠https://x.com/superlinear_fm⁠⁠LinkedIn: ⁠⁠https://www.linkedin.com/company/superlinearfm⁠⁠Brandon Kase: ⁠⁠https://x.com/bkase_⁠⁠Christine Yip: ⁠⁠https://x.com/christinetyip⁠Music licensed through Soundstripe. Code: O2UT6CCEQNH4KT0D

GPT-6 Astra and Fable 5.1 are changing how we start new software projects:Less harness engineering, less process, less prescribing. More deliberate steering on the few decisions that compound through the whole project.We walk through the Codex conversation of an actual project that we built: where we applied engineering judgment, what we left to the model, and how we planned verification so the agent could detect mistakes and keep building autonomously. From the first ramble to the stack and architecture decisions, this is our Fall 2026 workflow for starting a project with coding agents.Resources mentionedMatt Pocock — Grill with Docs The grilling skill Brandon uses to have the coding agent interview him, sharpen decisions, and build shared context before implementation. Grill with Docs Kit Langton Effect educator and practitioner whose work Brandon references for Effect patterns and tooling. https://kitlangton.com/ https://x.com/kitlangton Effect The TypeScript framework Brandon uses for typed errors, dependencies, concurrency, and application logic. Effect documentation Effect Solutions Practical patterns and reference material for building applications with Effect. https://www.effect.solutions/ Effect Institute Interactive resources for learning Effect. https://effect.institute/ Alchemy TypeScript infrastructure-as-code built around Effect, which Brandon uses to make infrastructure part of the codebase the agent can inspect and modify. Alchemy Lexi Lambda — “Parse, don’t validate” The type-driven design article behind the “make impossible states unrepresentable” principle discussed in the episode. Parse, don’t validate Rich Sutton — “The Bitter Lesson” The essay behind the “Bitter Lesson” idea we refer to when discussing why more capable models can make some bespoke scaffolding less useful. The Bitter Lesson / Rich Sutton’s publicationsSuperlinear is a podcast about emerging practices for building with AI.Hosted by Brandon Kase and Christine Yip.Discord: https://superlinear.fm/discordHomepage: ⁠⁠https://superlinear.fm⁠⁠Twitter: ⁠⁠https://x.com/superlinear_fm⁠⁠LinkedIn: ⁠⁠https://www.linkedin.com/company/superlinearfm⁠⁠Brandon Kase: ⁠⁠https://x.com/bkase_⁠⁠Christine Yip: ⁠⁠https://x.com/christinetyip⁠Music licensed through Soundstripe. Code: O2UT6CCEQNH4KT0D

Your project still carries decisions made before the latest model arrived: tests for removed features, slow checks, outdated code, and instructions that keep accumulating.A new model is a useful prompt to revisit those decisions. Give your agent a concrete problem to investigate, then review what its proposed changes would improve or put at risk.In this episode, we cover four maintenance tasks through our own projects: pruning low-value code and tests, turning recurring mistakes into lint rules, reassessing implementations, and updating agent instructions with official model guidance. Examples include an editor cleanup that removed about 6,000 net lines across 131 files, a custom Effect lint rule, and a solver investigation that cut round-trip time from 20 seconds to 1.5 seconds.Superlinear is a podcast about emerging practices for building with AI.Hosted by Brandon Kase and Christine Yip.Discord: https://superlinear.fm/discordHomepage: ⁠⁠https://superlinear.fm⁠⁠Twitter: ⁠⁠https://x.com/superlinear_fm⁠⁠LinkedIn: ⁠⁠https://www.linkedin.com/company/superlinearfm⁠⁠Brandon Kase: ⁠⁠https://x.com/bkase_⁠⁠Christine Yip: ⁠⁠https://x.com/christinetyip⁠Music licensed through Soundstripe. Code: O2UT6CCEQNH4KT0D

Your coding agent can inspect its own chat, but it cannot see you switching between browser research, documents, local files, and devices.Screen recordings can turn that invisible work into model context effectively with video native models like Gemini.In this episode, we show how Screenpipe exposed Christine acting as the context bridge around Codex, and Brandon's workflow for recording focused work sessions, compressing them to 720p at one frame per second, splitting them into 15-minute clips, and asking Gemini for a detailed play-by-play. We also extend the same idea to real phone test rigs, cameras, friction logs, and skills generated from recorded tasks.Superlinear is a podcast about emerging practices for building with AI.Hosted by Brandon Kase and Christine Yip.Discord: https://superlinear.fm/discordHomepage: ⁠⁠https://superlinear.fm⁠⁠Twitter: ⁠⁠https://x.com/superlinear_fm⁠⁠LinkedIn: ⁠⁠https://www.linkedin.com/company/superlinearfm⁠⁠Brandon Kase: ⁠⁠https://x.com/bkase_⁠⁠Christine Yip: ⁠⁠https://x.com/christinetyip⁠Music licensed through Soundstripe. Code: O2UT6CCEQNH4KT0D