rune-audit

Audits projects across eight dimensions and generates a weighted health score report.

1|Updated Mar 22, 2026
One-click install
npx skills add https://github.com/dangvu008/VietTruyen --skill rune-audit-dangvu008
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: rune-audit
Source: https://github.com/dangvu008/VietTruyen/tree/main/.agents/skills/rune-audit
Command: npx skills add https://github.com/dangvu008/VietTruyen --skill rune-audit-dangvu008

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams lack a systematic way to assess overall project health across security, dependencies, code quality, architecture, performance, infrastructure, and documentation, leading to hidden risks and unquantified technical debt. ## Core Features & Use Cases - 8-Dimension Health Audit: Runs phased audits covering dependencies, security, code quality, architecture, performance, infrastructure, documentation, and mesh analytics, delegating deep scans to specialist skills like sentinel and dependency-doctor. - Weighted Composite Scoring: Computes an overall health score using a weighted formula (Security 25%, Code Quality 20%, etc.) with grade thresholds from Excellent to Critical and a PASS/WARNING/FAIL verdict. - DX Review Mode: Evaluates developer experience using Addy Osmani's 8 principles, from time-to-hello-world to error recovery. - Use Case: Before a product launch, run a full audit to receive an AUDIT-REPORT.md with severity-ranked findings, file:line citations, prioritized actions, and a follow-up timeline. ## Quick Start Ask the agent to run a full project health audit and generate the AUDIT-REPORT.md with scores and prioritized fixes.

Frequently Asked Questions about rune-audit

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

FAQPage Schema
How do I run a full project health audit?

Invoke the audit with /rune audit or ask for a project health check. It runs eight phases covering dependencies, security, code quality, architecture, performance, infrastructure, documentation, and mesh analytics, then writes AUDIT-REPORT.md to the project root.

How is the overall project health score calculated?

The score uses a weighted formula: Security 25%, Code Quality 20%, Architecture 15%, Dependencies 15%, Performance 10%, Infrastructure 8%, and Documentation 7%. Scores map to grades from Excellent (90-100) down to Critical (0-39) with a PASS, WARNING, or FAIL verdict.

What is the DX review mode in a project audit?

DX review mode, triggered by /rune audit dx, evaluates developer experience using Addy Osmani's 8 principles, including time to hello world, setup friction, error message quality, and recovery from mistakes. It runs separately from the standard 8-phase health audit.

Does the audit support framework-specific checks?

Yes, it detects the stack during project discovery and applies targeted checks for React/Next.js, Node.js/Express, Python (Django/Flask/FastAPI), Go, Rust, and mobile frameworks like React Native and Flutter. Irrelevant framework checks are skipped entirely.

Why might an audit report mark a phase as incomplete?

A phase is marked incomplete when a delegated sub-skill such as sentinel or dependency-doctor times out, in which case the dimension receives an N/A score rather than fabricated findings. Every reported finding requires a specific file:line citation with over 80% confidence.