shadow-claw

Generate diverse hypotheses and structure evidence for multi-view fault investigations.

Updated Aug 31, 2024
One-click install
npx skills add https://github.com/iheCoder/Lib --skill shadow-claw
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: shadow-claw
Source: https://github.com/iheCoder/Lib/tree/main/skill/shadow_claw
Command: npx skills add https://github.com/iheCoder/Lib --skill shadow-claw

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Shadow Claw provides a structured, multi-perspective approach to diagnosing system faults by generating diverse hypotheses and organizing evidence into a traceable narrative that accelerates root-cause analysis.

Core Features & Use Cases

  • Multi-view hypothesis generation across execution paths, time/changes, resources, feedback loops, and data flows for comprehensive fault coverage.
  • Structured evidence chaining with explicit confidence levels, absence facts, and significance annotations to support evidence-driven decisions.
  • Investigation-cycle guidance with world expansion rules, hypothesis lifecycles, and convergence criteria to converge on root causes or escalate when needed.
  • Use Case: When a service experiences intermittent errors, Shadow Claw helps operators generate competing explanations, collect traces/logs, and validate hypotheses across multiple worlds to pinpoint the fault.

Quick Start

Initiate Shadow Claw for a reported incident by outlining the symptoms, then let it generate hypotheses and start evidence gathering.

Frequently Asked Questions about shadow-claw

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

FAQPage Schema
How do I structure root-cause analysis for complex service incidents using multiple investigation views?

Multi-view fault investigation coordinates competing hypotheses across execution paths, time changes, resources, feedback loops, and data flows to pinpoint root causes. It enforces a rigorous lifecycle with validation layers and convergence criteria to structure evidence for quick analysis.

What is the best way to validate competing fault hypotheses during a system outage?

Validating competing fault hypotheses requires structured evidence chaining with explicit confidence levels, absence facts, and significance annotations. This approach uses validation layers and convergence criteria to distinguish explanations and ensure accurate root-cause identification during an outage.

How do I trace intermittent service errors across data paths and execution flows?

Tracing intermittent service errors across data paths involves generating competing explanations and collecting logs to validate hypotheses across multiple worlds. Analysts examine execution paths, resource usage, and feedback loops to pinpoint the fault through controlled world expansion.

Can I use a structured investigation lifecycle for diagnosing network and data path faults?

A structured investigation lifecycle applies to diagnosing network and data path faults by guiding analysts through hypothesis generation, validation layers, and convergence criteria. This multi-perspective approach organizes evidence to accelerate root-cause analysis across complex services.

When should I use controlled world expansion during a multi-view fault investigation?

Controlled world expansion is used during a multi-view fault investigation when initial evidence is insufficient to converge on a root cause. It enforces investigation-cycle guidance to systematically expand execution paths and data flows, preventing premature conclusions.

How do I generate actionable reports from evidence captured during root-cause analysis?

Actionable reports from root-cause analysis are generated by enforcing a rigorous investigation lifecycle with five panels, including hypothesis lifecycle and convergence criteria. This structures captured evidence, absence checks, and significance annotations into a traceable narrative for quick decisions.