fable-thinking-ru

Enforce evidence-based verification and root-cause analysis for agentic coding tasks.

1|Updated Jul 3, 2026
One-click install
npx skills add https://github.com/uladzemer/fable-thinking --skill fable-thinking-ru
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: fable-thinking-ru
Source: https://github.com/uladzemer/fable-thinking/tree/main/ru
Command: npx skills add https://github.com/uladzemer/fable-thinking --skill fable-thinking-ru

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This skill addresses agentic failures in smaller models (Opus, Sonnet, Haiku) such as premature completion, symptom patching, and goal drift by enforcing a structured, verifiable operational discipline.

Core Features & Use Cases

  • Verification-First Workflow: Enforces the Proof-of-Work principle where no task is marked complete without external evidence (logs, test output, code reads).
  • Root-Cause Protocol: Implements a Two-Strikes rule to prevent cascading patches and forces a deep-dive debugging session when hypotheses fail.
  • Use Case: Use this when working on complex refactoring or debugging tasks on non-frontier models to ensure the agent maintains focus, verifies every assumption, and avoids "hallucinated" success.

Quick Start

Load the fable-thinking-ru skill at the start of your session to enforce rigorous self-verification and goal-anchoring procedures.

Frequently Asked Questions about fable-thinking-ru

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

FAQPage Schema
How do I prevent agentic coding loops from hallucinating success on complex debugging tasks?

To prevent hallucinated success in agentic coding, enforce a verification-first workflow that requires external evidence like logs or test output before marking any task complete. This structured discipline stops premature completion and forces iterative root-cause analysis.

Why does my AI coding agent apply cascading symptom patches instead of fixing the root cause?

Cascading symptom patches occur due to model drift in mid-tier models. Applying a Two-Strikes root-cause protocol stops this behavior by forcing a deep-dive debugging session when initial hypotheses fail, ensuring verifiable reasoning and goal-anchoring.

How do I maintain goal anchoring during multi-file refactoring with smaller AI models?

Maintain goal anchoring during multi-file refactoring by loading a strict operational discipline skill at the start of your session. This enforces external evidence-based verification and iterative procedures to prevent goal drift in mid-tier models like Sonnet or Haiku.

When do I need strict verification protocols for high-risk code modifications?

You need strict verification protocols for high-risk code modifications when using non-frontier models where model drift is a primary failure mode. This ensures every assumption is verified against external evidence, preventing symptom patching and premature completion.

Does this verification discipline work with mid-tier models like Opus and Haiku?

Yes, this verification discipline is specifically designed for mid-tier models including Opus, Sonnet, and Haiku. It enforces operational discipline and verifiable reasoning protocols to mitigate the specific agentic failure modes these smaller models exhibit.