EvanFlow: TDD

Trust Score 87/100

A vertical-slice TDD workflow: write one test, implement minimally, refactor, and repeat — focusing on behavior through public interfaces rather than implementa

triggers:test-driven developmenttddwrite testsred green refactorvertical sliceevanflow

GitHub SKILL.md

What it does

EvanFlow: TDD prescribes a disciplined test-driven development cycle focused on vertical slices. Each cycle is RED (write a failing test), GREEN (implement minimal code to pass), and REFACTOR (clean while tests are fresh). The skill documents vocabulary, anti-patterns, and a per-cycle checklist to enforce behavior-driven assertions and real-service integration where possible.

When to use it

Use it for developing production features, bug fixes, or refactors where behavioral guarantees matter. It's intended for backend services, business logic, and frontend code with testable behavior — avoid for throwaway prototypes or generated code. Use when you want tight feedback, safer refactors, and tests that describe observable behavior rather than internals.

What's included

Scripts: none included in the skill (has_scripts = false), but the repo contains meta-skills and related execution flows.
References: rich guidance and checklists embedded in the SKILL.md to aid disciplined cycles and assertion correctness.
Instructions: vocabulary, hard rules (never skip REFACTOR, vertical slices only), tracer bullet approach, and per-cycle refactor checklist.

Compatible agents

Suited to developer-assistant frameworks that can run tests and interact with repositories or CI (Claude Code, Codex-like agents, or CLI-capable tooling).

Audit Summary

A vertical-slice TDD workflow skill with no bundled scripts. Well-structured instructions emphasizing behavior testing over implementation testing, with an excellent assertion-correctness warning section. Clean security profile — pure instruction text with no executable code, network calls, or credentials.

Watch Out

Relies on evanflow meta-skill vocabulary defined externally
No scripts to automate anything — purely instructional

Notes

The assertion-correctness warning section (citing HumanEval research on LLM-generated test assertions) is a standout feature. The anti-pattern guidance on horizontal vs vertical slicing is practical and well-articulated.

Information

Repository: evanflow
Stars: 411

Trust Score

Overall87

Security95

Code Quality80

Architecture78

Usefulness72

Related Skills

Development Worktree

Create an isolated git worktree for feature work, auto-run project setup, and verify a clean test baseline before development.

ds-fix — data-science mid-analysis fixer

Orchestrates diagnosis and targeted fixes mid-analysis: diagnose root cause, apply fixes with output-first verification, and update project learnings.

Full Stack Builder

End-to-end builder that scaffolds, implements, tests, and optionally deploys web and API applications from a natural-language specification.

Claw Bench

Benchmarking skill that guides an agent through a structured suite of capability tests and reporting steps for leaderboard submission.

Visualization Best Practices

Provides step-by-step guidance and production-ready patterns for creating clear, accurate, and effective data visualizations in analytics workflows.

Make Scenario Creator

Assist with creating and validating 'make' scenario creator workflows for business automation: generate configurations, produce code, and recommend best practic

Ideogram Core Workflow — Edit, Remix, Upscale, Describe, Reframe

Image editing and image-to-image workflows for Ideogram: inpainting (Magic Fill), style remix, upscaling, image description, and reframing for new aspect ratios

Zero Tech Debt

Refactor features toward the intended architecture: remove legacy cruft and simplify code to reduce long-term operational cost.

Back to Skills