Codvo Internal L1 — Agent Foundations Practitioner

Purpose

Establishes common literacy. A certified practitioner understands what agents are, when they are appropriate, what makes them risky, and how to map an agent workflow into control layers.

Competencies — practitioners must be able to

• Explain the difference between chatbot, copilot, workflow automation, and AI agent
• Identify appropriate and inappropriate agent use cases
• Describe basic agent patterns: tool use, routing, planning, evaluation, human-in-the-loop
• Explain the seven control layers for production agents
• Identify common agent risks: hallucination, tool misuse, prompt injection, data leakage, excessive autonomy, weak auditability
• Map a business workflow into agent steps, data sources, tools, approvals, risks, and expected outcomes
• Communicate agent value and risk in CIO-friendly language

Learning modules

1. 1Agent fundamentals and definitions
2. 2Agent patterns and workflow design
3. 3Use case selection and value scoring
4. 4Control-layer architecture overview
5. 5Basic threat and risk awareness
6. 6Industry scenario walkthrough

Required deliverables

• Agent Opportunity Canvas
• Control-Layer Map v1
• Basic Risk Checklist
• Business Value Statement
• 10-minute walkthrough recording or live presentation

Practical capstone

Select one realistic workflow and produce a one-page agent opportunity canvas covering business problem, proposed agent role, users and stakeholders, data sources, tools and systems, human approval points, risks and controls, success metrics.

Assessment method

Artifact review and short presentation. Weighted: use case clarity (20%), classification (15%), control-layer coverage (25%), risk identification (20%), communication quality (20%).

Pass criteria

Score ≥75% overall and no zero-score on risk identification or control-layer coverage.