Does Flow Prime modify my source code?

No. Flow Prime is designed to only modify or create configuration files, documentation (like CLAUDE.md), and scripts. It never alters your application logic and never commits changes automatically.

What is Agent Readiness?

Agent Readiness measures how effectively an AI agent can navigate, build, and test your codebase. Flow Prime fixes the environment issues that cause agents to fail, such as missing documentation or broken test scripts.

What are the Eight Pillars of assessment?

The assessment covers Style & Validation, Build System, Testing, Documentation, Dev Environment (Agent Readiness), and Observability, Security, and Workflow (Production Readiness).

How long does a Flow Prime scan take?

Thanks to its parallel scout architecture, a full assessment typically completes in 15 to 20 seconds, even for complex repositories.

Flow Prime Codebase Readiness

Name: Flow Prime Codebase Readiness
Author: gmickel

bygmickel

•

472

•

Security & Testing

Assesses and optimizes your codebase for AI agent autonomy and production quality through a multi-pillar health scan.

Flow Prime is a comprehensive assessment tool designed to bridge the gap between human-centric codebases and agent-ready environments. By scanning your project across 8 pillars and 48 specific criteria, it identifies why AI agents might struggle—such as missing CLAUDE.md files, undocumented environment variables, or broken test commands. The skill uses 9 parallel AI scouts for rapid analysis, providing a maturity score and offering interactive remediation for agent-readiness gaps while reporting on production-level concerns like security and observability. It ensures your environment is optimized for autonomous operation, reducing the 'friction cycles' that waste time and compute.

Key Features

01Maturity Scoring: Categorizes projects into 5 levels, from Minimal to Autonomous, with targeted benchmarks.

02Automated Remediation: Proposes and applies fixes for environment, documentation, and tooling gaps.

038-Pillar Health Scan: Evaluates 48 criteria across agent readiness and production standards.

04472 GitHub stars

05Command Verification: Actively tests and verifies that build and test commands actually work in the current environment.

06Parallel AI Scouts: Deploys 9 concurrent Haiku scouts to analyze codebase health in under 20 seconds.

Use Cases

01Standardizing developer experience (DX) across teams to ensure fast feedback loops and reliable builds.

02Onboarding a legacy repository for high-efficiency use with Claude Code or autonomous agents.

03Auditing project health to identify missing documentation, security gaps, and CI/CD weaknesses.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add gmickel/gmickel-claude-marketplace flow-next-prime

For use in Claude.ai and ChatGPT

Download Skill