How does Argus ensure the accuracy of its findings?

Argus follows a strict policy requiring every finding to carry verifiable evidence and mandates a certification step using 'context_certify' before final report publication.

When should I use the 'argus on' command?

You should use 'argus on' when you need to deep-dive into a specific subsystem, analyze a complex block of code, or create a comprehensive map of a project's architecture.

What is the primary purpose of the Argus skill?

Argus is designed to provide a high-confidence diagnostic view of specific code blocks or subsystems, identifying verifiable findings and opportunities for improvement.

What is the 'Observation Cycle' used by this skill?

The Observation Cycle is a 5-step process: building a scope envelope, executing an evidence sweep, classifying findings, certifying report quality, and recording the outcome.

Argus Subsystem Diagnostic

Name: Argus Subsystem Diagnostic
Author: murillodutt

bymurillodutt

0•

Analytics & Monitoring

Performs deep, evidence-based diagnostic examinations of software subsystems to identify architectural improvements and monitoring requirements.

Argus is a specialized diagnostic skill for Claude Code designed to provide high-confidence analysis of specific code blocks or complex subsystems. Built on the Cellm context engineering framework, it utilizes a structured 'Observation Cycle' that includes preflight scope definition, multi-lens evidence gathering, and a mandatory certification process. By classifying findings into actionable categories—such as immediate construction opportunities or monitor-only items—Argus ensures that developers receive verifiable insights rather than mere speculation, making it an essential tool for mapping legacy code and validating architectural integrity.

Key Features

01Structured 'Observation Cycle' for rigorous analysis

02Automated certification and outcome recording for continuous learning

030 GitHub stars

04Classification of insights into operate, construct, and monitor categories

05Evidence-based finding verification using local system tools

06High-confidence subsystem mapping and diagnostic views

Use Cases

01Generating detailed technical reports and observability findings for subsystem health.

02Performing evidence-based audits of specific code blocks before significant refactoring.

03Mapping out complex legacy subsystems to understand hidden dependencies and logic flow.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add murillodutt/cellm argus

For use in Claude.ai and ChatGPT