skeptic-clean

Run auditable data cleaning cycles within the Skeptic framework.

Updated Apr 9, 2026
One-click install
npx skills add https://github.com/Filivignaga/skeptic --skill skeptic-clean
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: skeptic-clean
Source: https://github.com/Filivignaga/skeptic/tree/main/codex/skeptic-clean
Command: npx skills add https://github.com/Filivignaga/skeptic --skill skeptic-clean

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Skeptic auditable data cleaning: Use after formulate and protocol to build an auditable cleaning pipeline under protocol-defined data visibility, without widening the claim boundary or assuming predictive workflow defaults. Third stage of Skeptic. Use when Codex should run the Skeptic clean stage as a standalone skill, including requests like skeptic clean --auto to run this stage with autonomous cycle execution.

Core Features & Use Cases

  • Auditable, reproducible cleaning pipeline aligned with formulate and protocol.
  • Standalone execution via skeptic clean or autonomous cycles using --auto.
  • Produces canonical YAML artifacts, compact decision ledger, and cycle evidence per cycle.

Quick Start

Run skeptic clean in your project context to execute the auditing cleaning cycles, or use --auto to run all cycles autonomously.

Frequently Asked Questions about skeptic-clean

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

FAQPage Schema
What is auditable data cleaning and how does it enforce reproducibility?▼

Auditable data cleaning enforces protocol-driven data visibility and prevents claim widening. It produces canonical YAML artifacts and a compact decision ledger to support full reproducibility and auditability of the pipeline.

How do I automate reproducible data cleaning cycles in an existing project?▼

Run skeptic clean in your project context to execute the auditing cleaning cycles. Use the --auto flag to run this stage with autonomous cycle execution and generate cycle evidence automatically.

Do I need to run formulate and protocol stages before standalone data cleaning?▼

Yes, you should run formulate and protocol beforehand. This skill operates as the third stage to build an auditable cleaning pipeline under protocol-defined data visibility without widening the claim boundary or assuming predictive workflow defaults.

How does protocol-driven data cleaning prevent claim widening during analysis?▼

Protocol-driven data cleaning restricts data visibility to protocol-defined boundaries. By producing canonical artifacts and a decision ledger, it ensures the cleaning process does not widen the claim boundary or assume predictive workflow defaults.

What artifacts does an auditable data cleaning pipeline produce for auditability?▼

The cleaning pipeline produces canonical YAML artifacts, a compact decision ledger, and cycle evidence per cycle. These outputs support reproducibility and ensure the data cleaning decisions are fully auditable.