Judgment Infrastructure for Human-Led AI

Portable agent skills by Greg Lichtenthal for people who want AI to strengthen human creativity, judgment, agency, and evidence-seeking action.

Start Here

Messy context?

Use The Briefing Room to turn notes, transcripts, research, and fragments into a brief.

A claim or plan?

Use Ground Truth to pressure-test the thinking before you act on it.

A consequential decision?

Use The Quorum to deliberate trade-offs through multiple expert lenses.

An idea, claim, or decision?

Use Test Drive to turn it into the smallest credible test.

The Judgment Loop

You do not need the full loop every time. Start with the part that matches what you have: messy context, a claim to test, a consequential decision, or an idea that needs evidence.

Mess raw context, uncertainty, fragments
Brief The Briefing Room makes it legible
Critique Ground Truth challenges the thinking
Deliberate The Quorum surfaces trade-offs
Test Test Drive creates evidence

Use Cases

Product strategy

Use the loop for customer feedback, roadmap bets, positioning, and prioritization. Open use case.

Founder decisions

Pressure-test pivots, senior hires, fundraising narratives, and strategic bets. Open use case.

Research synthesis

Turn papers, interviews, notes, and evidence dumps into structured judgment. Open use case.

Career decisions

Clarify offers, role changes, trade-offs, and small experiments without outsourcing judgment. Open use case.

Ground Truth social preview
Claude Codex Critical thinking

Ground Truth

Calibrated honesty and anti-sycophancy for plans, decisions, reviews, and ideas.

Frequently Asked Questions

What is judgment infrastructure for human-led AI?

It is a set of portable AI skills that help people structure messy information, challenge their own reasoning, and make better decisions without outsourcing human judgment.

What is an agent skill?

An agent skill is a portable set of instructions, examples, and metadata that helps an AI agent perform a specialized workflow more reliably than a one-off prompt.

What is evidence design?

Evidence design means identifying what kind of signal would make an idea more or less trustworthy, then creating the smallest useful test to gather that signal.

When should I use The Briefing Room, Ground Truth, The Quorum, or Test Drive?

Use The Briefing Room when the context is messy, Ground Truth when you want honest critique on a plan or claim, The Quorum when you need multiple expert lenses on a consequential decision, and Test Drive when you need evidence before committing.

How do I install these skills in Codex or Claude?

Each skill has a GitHub repository, a release download for the packaged skill, and installation instructions for Claude and Codex. The catalog links directly to all four install files.