agent-architecture-audit

Audit agent codebases against 12 production primitives and generate an upgrade roadmap.

4|1|Updated Apr 11, 2026
One-click install
npx skills add https://github.com/m2ai-portfolio/m2ai-skills-pack --skill agent-architecture-audit
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-architecture-audit
Source: https://github.com/m2ai-portfolio/m2ai-skills-pack/tree/main/skills/agent-architecture-audit
Command: npx skills add https://github.com/m2ai-portfolio/m2ai-skills-pack --skill agent-architecture-audit

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Identifies gaps in agent infrastructure by auditing an agent codebase against 12 production primitives and delivering a structured upgrade plan.

Core Features & Use Cases

  • 12-primitive readiness assessment across permission model, token budget, crash recovery, tool assembly, streaming events, state machine, provenance, stop reasons, boot sequence, verification harness, memory decay, and provenance-aware context.
  • Prioritized upgrade path with concrete actions to improve production readiness for agent architectures.
  • Works on standard agent codebases and directories to surface gaps and remediation steps.

Quick Start

Run the audit against your agent codebase to generate a prioritized upgrade plan.

Frequently Asked Questions about agent-architecture-audit

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

FAQPage Schema
How do I audit my agent codebase for production readiness?

Audit agent production readiness by scanning your codebase against 12 infrastructure primitives. The assessment generates a present, partial, or missing rating for each primitive, an overall score, and a prioritized upgrade roadmap.

What are the 12 production primitives for agent architecture?

The 12 production primitives for agent architecture include permission model, token budget, crash recovery, tool assembly, streaming events, state machine, provenance, stop reasons, boot sequence, verification harness, memory decay, and provenance-aware context.

How do I identify infrastructure gaps in my AI agent before deployment?

Identify agent infrastructure gaps by running a gap-analysis audit on your repository directory. It surfaces missing or partial production components and delivers a structured scorecard with concrete remediation steps.

Can I audit a specific directory path within my agent repository?

Yes, you can audit user-specified directory paths containing agent code. The audit applies the 12-primitive scoring model to the targeted path to surface architecture gaps and generate a targeted upgrade plan.

What is the best way to plan an upgrade path for an agent architecture?

Plan an agent architecture upgrade path by using a phased scoring model that rates each production primitive. It produces a prioritized roadmap with concrete actions to improve production readiness.

Why does my agent crash and how do I assess its recovery mechanisms?

Assess agent crash recovery by auditing the codebase against the crash recovery primitive. The scorecard rates crash recovery as present, partial, or missing and provides remediation steps within the upgrade roadmap.