fabuleux

What problem does it solve?

Most AI agents produce bloated, unverified output, pad content with unnecessary structure, and give overly optimistic status reports that hide real problems, leading to deliverables that do not actually work as expected or contain hidden defects.

Core Features & Use Cases

• Task-type routing: Automatically classifies tasks into the appropriate workflow (artefact development, prose writing, analysis, auditing, or simple one-shot queries) to apply the right discipline for each task type.
• Visual self-evaluation for artefacts: Mandates screenshot capture and vision-based defect checking for visual deliverables like web pages, code, and documents to catch layout, alignment, and functionality issues before delivery.
• Prose subtraction pass: Enforces a mandatory ~20% content cut for written work to eliminate fluff, padding, and unnecessary structure, resulting in more natural, concise text.
• Honest audit workflow: Provides a 3-step process for auditing existing deliverables without modification, then applying targeted corrections with full re-verification and honest status reporting.
• Use cases: Building web pages, code, documents, or data pipelines that need to function correctly; auditing existing deliverables for hidden issues; writing concise, natural prose; any multi-step task where honest status reporting and verified results are required.

Quick Start

Ask Claude to activate the fabuleux skill to build a responsive landing page, verify it renders correctly on desktop and mobile, fix all visual defects, and provide an honest report of the final deliverable state.