doubt-driven-development
Subjects every non-trivial decision to a fresh-context adversarial review before it stands. Use when you want every assumption cross-examined before proceeding, when stress-testing a plan for hidden failure modes, when correctness matters more than speed, when working in unfamiliar code, when stakes are high (production auth, security-sensitive logic, a high-stakes migration, irreversible operations), or any time a…
- installs 8w
- 18,694
- 30-day movement
- starts with the next reading
- Related entries
- 6
- Connections
- 0
An agent skill that forces every non-trivial decision through a fresh-context adversarial review before it is accepted. It defines a five-step loop — claim, extract, doubt, reconcile, stop — where a reviewer with no shared context cross-examines the decision, with optional cross-model escalation.
Long agent sessions quietly turn assumptions into facts; this catches them while verification is still cheap.
Use it to
- Stress-test a plan for hidden failure modes
- Review security-sensitive or irreversible changes
- Verify decisions in unfamiliar codebases
- Cross-examine assumptions before a high-stakes migration
- Escalate to a different model to escape shared blind spots
For Developers using AI coding agents on high-stakes work
- Host repository
- addyosmani/agent-skills
- Installs, lifetime
- 29k
- Installs, 8 weeks
- 19k
- Host stars
- 96k
- Host language
- JavaScript