What Happens When You Hit Claude’s Limits? | Sam Hesson (Meta AI)

90 snips

Feb 25, 2026

Sam Hesson, a technical founder turned Meta AI incubations engineer who built agentic CI/CD pipelines, tells wild stories and explains his architecture. He covers his $50k token-abundance mindset. Short, punchy dives into converting Ray-Bans conversations to PRDs, running parallel competitive agents, rubric-based LLM judges for self-healing PRs, and prepping codebases for reliable agent testing.

Ask episode

AI Snips

Chapters

Transcript

Episode notes

INSIGHT

Token Abundance Mindset Unlocks New Workflows

Sam calls a Token Abundance Mindset the shift to treating tokens as cheap experiment budget to explore agentic dev at scale.
That mindset enabled running long multi-turn runs and thinking in parallelized agent workflows rather than single prompts.

ADVICE

Convert Walk Conversations Into Ready PRDs

Save voice-walk conversations as markdown and ingest them into your repo to generate PRDs and tech specs automatically.
Sam runs scripts that convert those files into implementation-aware prompts using codebase-derived rules before pushing to GitHub.

ADVICE

Make Playwright Tests Part Of Agent Tasks

Add Playwright end-to-end scripts to capture UI behavior and auto-generate unit tests during agent development.
Sam treats Playwright tests as feature development guards and asks agents to create unit tests every time they change UI.

Get the Snipd Podcast app to discover more snips from this episode

Get the app

Description:

Sam Hesson — technical founder and newly minted member of Meta's AI incubation team — joins the AI Tinkerers One-Shot podcast for a deep-dive into the custom CI/CD pipeline he built to orchestrate parallel AI agents at scale. Sam became notorious for spending $50,000 on tokens in just two months while pushing the limits of what agentic development systems can do. In this episode, he shares his complete architecture.

Sam walks through his "Token Abundance Mindset" and explains why moving beyond single-turn prompting is the key to unlocking a fully automated, multi-agent development workflow. From structuring your codebase for AI readiness to running adversarial LLM judges that self-heal pull requests, this is one of the most technically advanced conversations we've had on the show.

Topics covered in this episode:

The Darwinian CI/CD — running multiple parallel agents in competition to produce the best pull request
LLM Judge — rubric-based quality control and adversarial agents for self-healing PRs
AI Dev Readiness — DB seeding, snapshots, and codebase prep for reliable agent testing
The Walk Flow — converting high-level conversations (via Meta Ray-Bans) into structured PRDs and Tech Specs
DevPlan — generating detailed, unambiguous coding prompts that eliminate guesswork for agents

Timestamps:

00:00 Introduction: Sam Hesson & Claude Maxing

02:17 The Alpha of 10x Agentic Systems

03:30 The $50k Token Spend & Abundance Mindset

07:25 The Walk Flow: Ray-Bans to PRD/Tech Spec

11:17 AI Adherence to Implementation Patterns

15:21 Playwright & AI Codebase Readiness

17:35 The Darwinian CI/CD: Parallel Agents

21:47 LLM Judge & Rubric-Based Self-Healing

31:03 Bruno API Client: Code-Like Specs

32:47 Self-Healing PRs & Agentic Graph RAG

38:40 DevPlan: Turnkey Structured Prompting

58:43 The Future of SDLC & AI Agent Adoption

Resources:

Sam Hesson on LinkedIn: linkedin.com/in/samhessenauer
Nanome.ai (Sam's former company)
AI Tinkerers: aitinkerers.org
One-Shot Podcast: aitinkerers.org/podcast

AI Tinkerers One-Shot is a podcast for builders and innovators defining the future of AI — going under the hood with the people actually shipping it.