red/green TDD

Overview

Red/green TDD is a test-driven development pattern adapted for coding agents. In this workflow, an agent is prompted to write tests first, run them to confirm they fail (the “red” phase), implement the code, and then rerun the tests to confirm they pass (the “green” phase). In the context of agentic software development, this pattern acts as a reliability scaffold: it gives the model an explicit loop for validating intent, reducing the chances of silent errors or partially correct implementations.

For AI Product Managers, red/green TDD matters because it turns coding agents from unpredictable generators into more governable execution systems. Instead of evaluating output only after a large block of code is produced, teams can structure agent work around short validation cycles. This is especially useful when shipping internal tools, prototypes, integrations, and product experiments with agents such as Claude Code or GPT-5.4, where speed is valuable but correctness still needs guardrails.

Key Developments

• 2026-02-23: Simon Willison’s Agentic Engineering Patterns was highlighted as a hub for patterns used to build and operate agentic systems, including red/green TDD for coding agents.
• 2026-04-03: A practical description of the pattern was shared: prompting an agent with “red/green TDD” instructs it to write tests first, run them to verify failure, implement the code, and rerun tests to verify success.
• 2026-04-04: Lenny Rachitsky shared Simon Willison’s view that late 2025 marked an inflection point for AI coding, with autonomous coding agents, benchmark progress, and workflows like red/green TDD plus “thin templates” improving reliability and productivity.

Relevance to AI PMs

1. Improve reliability of agent-built features: AI PMs can require red/green TDD for agent-assisted coding tasks so outputs are validated against explicit tests before being accepted into a codebase.
2. Operationalize faster experimentation: For prototypes, internal tooling, and product spikes, this pattern helps teams move quickly while preserving a lightweight quality bar through automated failure/success checks.
3. Standardize agent workflows across teams: PMs can turn red/green TDD into a delivery norm or prompt template, making outcomes more predictable when multiple engineers or agents contribute to the same product surface.

Related

• Simon Willison: Closely associated with popularizing agentic engineering patterns, including red/green TDD for coding agents.
• Lenny Rachitsky: Amplified the idea through newsletter and social distribution, helping bring the pattern to a broader product and tech audience.
• Claude Code: An example of a coding agent environment where red/green TDD can be used to structure implementation loops.
• GPT-5.4: Referenced as part of the wave of stronger coding agents that made patterns like red/green TDD more practical and productive.
• Agentic Engineering Patterns: The broader framework or collection of practices in which red/green TDD appears as a specific implementation pattern.
• Test-driven development: The foundational software practice that red/green TDD adapts for agentic coding workflows.