Dialectic Loop

Installation
SKILL.md

Dialectic Loop Skill

Validate and refine an empirical claim or model by running C.S. Peirce's inquiry cycle — abduction → deduction → induction — across three roles that debate and update a hypothesis against real data.

Overview

A Dialectic Loop improves the quality of a claim about reality (a trend, a pattern, a characterization, an empirical model) by:

  1. Deriving falsifiable predictions from a hypothesis (deductive role)
  2. Testing those predictions against a real corpus — surfacing supporting instances and actively hunting counterexamples (inductive role)
  3. Arbitrating the gap between prediction and evidence, then updating the hypothesis (arbiter role)
  4. Iterating until the hypothesis stabilizes (converges)

The output is not approve/reject and not a root cause — it is a refined hypothesis (H′) with an evidence-backed confidence level and a record of what the original framing got wrong.

Key feature: In codex mode, the inductive role is assigned to Codex so the empirical test is performed by a different model than the one that authored the hypothesis. This independence is the point — it is what catches the hypothesis author's confirmation bias.

Comparison with sibling skills

Installs
GitHub Stars
3
First Seen
Dialectic Loop — masup9/codex-collab