one step labs
Claude Plugin onesteplabs.com
Methodology

AI Process
Assessment

A Claude plugin that runs a full consulting methodology: finding, scoring, and roadmapping automation opportunities from scoping through to client delivery.

11 phases · 2 gates
How it works
Phases 1–3
Understand
  • Scope the engagement & goals
  • Map organizational context
  • Inventory tech, data & systems
Checkpoint 1
Phases 4–6
Discover & Score
  • Interview & map every process
  • Identify & type AI opportunities
  • Score across 6 dimensions
Compliance Gate · Checkpoint 2
Phases 7–11
Sequence & Deliver
  • Prioritize 3-wave roadmap
  • Package Wave 1 use case briefs
  • Business case with cost estimates
  • Single-scroll client deliverable
Deliverable Gate · Checkpoint 3
Stakeholder validation at
Scope & Context Baseline Processes Portfolio & Roadmap

What makes it different

Four defining principles
Foundation
AI-first
Built for AI agents, not retrofitted from human consulting workflows. Each phase is a self-contained task with defined inputs, outputs, and quality checks built in.
Modularity
Skills
Each phase is powered by a dedicated skill: a focused capability with its own instructions, logic, and output format. Skills run on demand and can be called in sequence or on their own.
Parallelism
Agents
Separate, independent agents run in parallel to review, score, and verify work at each phase. A reviewer checks every opportunity before it can advance. The main agent never grades its own work.
Integrity
Self-fixing
Quality gates prevent work from advancing until it passes review. Compliance and quality checks run at key points in the process. Every figure in the final deliverable traces back to a source file.