analyze

Analyzes architecture, bugs, dependencies across files, logs, and runtime traces to identify root causes.

Updated Apr 3, 2026

Search Tags:#architecture #investigation #evidence #analyze #root-cause #hypotheses

Ready to run

npx skills add https://github.com/pextacy/stellar --skill analyze-pextacy

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: analyze
Source: https://github.com/pextacy/stellar/tree/main/skills/analyze
Command: npx skills add https://github.com/pextacy/stellar --skill analyze-pextacy

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Helps analysts and engineers diagnose ambiguous, causality-driven questions by structuring observations, testing hypotheses, and weighing evidence across multiple sources to identify root causes.

Core Features & Use Cases

Structured reasoning: Observation, Hypotheses, Evidence For, Evidence Against, Current Best Explanation, Critical Unknown, Discriminating Probe.
Cross-file analysis: Integrates findings from code, configs, logs, and artifacts to support robust conclusions.
Outputs & guidance: Produces Observed Result, Ranked Hypotheses, Evidence Summary, Evidence Against / Gaps, Most Likely Explanation, Critical Unknown, and Recommended Next Step.

Quick Start

Provide a structured investigation by listing observations, formulating hypotheses, gathering evidence, and stating the current best explanation.

Frequently Asked Questions about analyze

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

FAQPage Schema

How do I perform a root-cause analysis across scattered code, logs, and configs?▼

Root-cause analysis across scattered code, logs, and configs requires formalizing observations, ranking hypotheses, and weighing evidence. This Skill integrates cross-file findings to output a most likely explanation and a recommended next step.

What is the best way to structure an investigation for ambiguous software bugs?▼

Structuring an investigation for ambiguous bugs involves cataloging observations, formulating hypotheses, and mapping evidence for and against each. This approach yields ranked hypotheses and identifies critical unknowns for targeted probing.

When do I need evidence-driven hypothesis testing for architecture analysis?▼

Evidence-driven hypothesis testing is needed when architecture analysis requires cross-file reasoning over dispersed artifacts. It formalizes current best explanations by validating hypotheses against runtime traces and configuration evidence.

Can I use this approach to investigate dependency-related issues and runtime traces?▼

Yes, investigating dependency issues and runtime traces fits this approach by correlating dispersed evidence. It evaluates evidence for and against hypotheses to determine the most likely explanation and recommend a discriminating probe.

Why does my root-cause investigation fail to identify the actual bug?▼

Root-cause investigations fail when they lack structured evidence weighing and ignore critical unknowns. Formalizing hypotheses against cross-file evidence gaps ensures a discriminating probe is identified to validate the most likely explanation.

Does this root-cause analysis method require any specific dependencies or components?▼

No specific dependencies or components are required. The method applies structured reasoning to existing artifacts like code, configs, and logs to produce evidence summaries and recommended next steps.