doubt-driven-development

Subjects architectural and logic decisions to adversarial, fresh-context review via a five-step verification process.

Updated Jul 31, 2026
One-click install
npx skills add https://github.com/fsolla/teqo --skill doubt-driven-development-fsolla
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: doubt-driven-development
Source: https://github.com/fsolla/teqo/tree/main/.agents/skills/doubt-driven-development
Command: npx skills add https://github.com/fsolla/teqo --skill doubt-driven-development-fsolla

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill prevents costly errors in high-stakes development by subjecting non-trivial decisions to a fresh-context, adversarial review before they are finalized.

Core Features & Use Cases

  • Adversarial Scrutiny: Forces the AI to act as a skeptic, identifying hidden assumptions, edge cases, and failure modes in your logic.
  • Structured Verification: Uses a rigorous five-step process (Claim, Extract, Doubt, Reconcile, Stop) to ensure decisions are sound.
  • Use Case: Use this when implementing complex branching logic, security-sensitive code, or irreversible production changes to ensure correctness before committing.

Quick Start

Invoke the doubt-driven development skill to perform an adversarial review of the proposed architectural change for the new caching layer.

Frequently Asked Questions about doubt-driven-development

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
What is adversarial review in software architecture and when do I need it?

Adversarial review subjects non-trivial architectural and logic decisions to fresh-context scrutiny to identify hidden assumptions and failure modes. You need it for high-stakes development tasks including production deployments, security-sensitive logic, and irreversible data operations before finalizing changes.

How do I verify complex branching logic before a production deployment?

Verify complex branching logic by applying a structured five-step process: Claim, Extract, Doubt, Reconcile, and Stop. This isolates artifacts from reasoning to ensure objective, issue-focused evaluation, identifying edge cases and failure modes in your logic before committing.

Can I use adversarial review for security-sensitive code and irreversible data operations?

Yes, adversarial review is designed for security-sensitive code and irreversible data operations. It forces a skeptical evaluation of your proposed logic to find hidden assumptions and failure modes, ensuring correctness before you commit to high-stakes, irreversible changes.

What's the best way to identify hidden assumptions in architectural decisions?

The best way to identify hidden assumptions is a fresh-context, adversarial review that isolates artifacts from reasoning. By forcing a skeptical perspective through a structured five-step verification process, you uncover edge cases and failure modes that standard reviews often miss.

When should I not use an adversarial review process for code verification?

You should not use adversarial review for trivial changes that lack architectural impact or complex logic. This rigorous five-step verification process targets non-trivial, high-stakes decisions where identifying hidden assumptions and failure modes justifies the added scrutiny.