launch-project-audit

Audit a GitHub codebase to classify feature build states and output evidence tables.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/CleanExpo/Pi-Dev-Ops --skill launch-project-audit
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: launch-project-audit
Source: https://github.com/CleanExpo/Pi-Dev-Ops/tree/main/skills/launch-project-audit
Command: npx skills add https://github.com/CleanExpo/Pi-Dev-Ops --skill launch-project-audit

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill exposes what is truly implemented versus what is only promised, catching the common hidden gap where code exists but nothing calls it.

Core Features & Use Cases

  • Build-state feature mapping: Classifies each feature as not-started, stubbed, built-disconnected, built-wired, or built-tested using concrete repository signals.
  • Evidence-cited wiring & production progress: Determines whether features are connected and estimates how far they’ve traveled through the pipeline (idea → in-progress → PR-open → merged → in-production).
  • Frontier escalation with cost control: Runs a cheap bulk pass first and escalates to a frontier model only for ambiguous cases, while remaining read-only.

Quick Start

Ask the agent to scan the repository to produce a build-state map that identifies finished, wired, tested, and built-disconnected features with evidence and suggested next steps.

Frequently Asked Questions about launch-project-audit

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

FAQPage Schema
What is a production readiness audit for a GitHub codebase?

Production readiness audits classify features as not-started, stubbed, built-disconnected, built-wired, or built-tested. They determine whether code is actually connected and estimate deployment progress from in-progress to in-production.

Can I check if my repository code is actually wired and tested versus stubbed?

Map pipeline staging to deployment evidence by tracking feature progress from idea through PR-open and merged to in-production. The audit assesses how far each feature has traveled through the deployment pipeline.

Does the codebase audit modify my repository files during the scan?

Yes, you can check if code is wired and tested by scanning repository signals. The audit classifies features as stubbed, built-disconnected, built-wired, or built-tested based on concrete evidence.

What is the best way to handle ambiguous feature states during a codebase audit?

No, the codebase audit stays strictly read-only and safety-bound. It analyzes the repository and saves the resulting build-state map and logged curator proposals to a local .harness/audits directory.

How do I generate an evidence-based report of what is built versus what is left?

The best way to handle ambiguous feature states is a frontier escalation with cost control. The audit runs a cheap bulk pass first and escalates only ambiguous cases to a frontier model.

How do I generate an evidence-based report of what is built versus what is left?

Generate an evidence-based report by scanning the codebase to produce a build-state map. It identifies finished, wired, tested, and built-disconnected features with evidence and suggested next steps saved to .harness/audits.