Drift Documentation
Test-driven development for AI workflows.
What Drift Does
Drift is a TDD framework for AI workflows. Define your standards first, validate against them, fix issues, and iterate - just like TDD for code.
TDD Workflow for AI Agent Projects
Define standards - Write validation rules in
.drift.yamlfor your expected project structureRun validation - Execute
driftto see what’s missing or broken (red phase)Fix issues - Create files, fix links, update configurations manually or with AI assistance (green phase)
Iterate - Re-run validation until all checks pass (refactor phase)
No built-in opinions - You define your team’s standards in .drift.yaml, then validate against them. Perfect for setting up new projects or enforcing consistency across teams.
Project Structure Validation (Primary Use)
Define rules in .drift.yaml to check:
Dependency health: Detect redundant transitive dependencies in commands, skills, and agents
Link integrity: Validate all file references and resource links in documentation
Completeness checks: Ensure skills, commands, and agents have required structure
Configuration validation: Verify agent tools format, frontmatter schema, and MCP permissions
Consistency validation: Detect contradictions between commands and project guidelines
Required files: Verify essential configuration files exist
See Configuration for rule examples and Quick Start Guide for starter rules.
Conversation Quality Analysis (Optional)
For teams wanting deeper insights, run drift with LLM-based rules to analyze AI agent conversations. Define custom conversation-level validation rules in .drift.yaml to detect patterns like:
User manually executing steps that slash commands automate
AI agent not using available skills for specialized tasks
Available automation or capabilities not being utilized
Conversations diverging from documented best practices
Requires LLM access and rules with type: prompt.