Founder Field Guides

Frameworks for the work that has to hold up.

Use these practical guides to pressure-test an operating model, AI decision path, production release, or critical automation before the weakness becomes expensive.

Operating model

The Multi-Role System Blueprint

Map roles, decisions, handoffs, boundaries, exceptions, proof, and automation before funding more screens.

Open the guide →
AI evidence

The AI Claim-Proof Protocol

Turn citations from decoration into a testable evidence contract for important AI-generated claims.

Open the guide →
AI decisions

The AI Rejection Audit

Inspect the decisions automation hides and turn structured disagreement into a measurable review loop.

Open the guide →
Workflow reliability

The Silent Success Audit

Trace a workflow from accepted request to a visible, recoverable outcome for the person waiting.

Open the guide →
Automation safety

The Retry-Safe Automation Playbook

Give one business intent a stable identity so retries do not duplicate an expensive action.

Open the guide →
Release proof

The Shipped-or-Not Release Audit

Demand evidence that work is tracked, reproducible, deployed, reachable, usable, and reversible.

Open the guide →
AI security

The Public AI Feature Threat Model

Review the browser, server, model, renderer, data, cost, and recovery boundaries before public traffic.

Open the guide →
Data integrity

The Active Workspace Integrity Audit

Keep reads, writes, billing, automation, and cached state bound to the workspace the user can see.

Open the guide →