trellis-check

Verifies code changes against specs, lint, type checks, tests, and cross-layer consistency.

Updated Apr 14, 2026
One-click install
npx skills add https://github.com/exhyy/face-crop --skill trellis-check-exhyy
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: trellis-check
Source: https://github.com/exhyy/face-crop/tree/main/.claude/skills/trellis-check
Command: npx skills add https://github.com/exhyy/face-crop --skill trellis-check-exhyy

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? After writing code, it is easy to miss spec violations, broken tests, inconsistent patterns across layers, or duplicated logic before committing. This Skill runs a structured quality verification pass over recently changed code so issues are caught and fixed before they reach the repository. ## Core Features & Use Cases - Spec Compliance Verification: Reads the applicable Trellis spec indexes and guideline files for each changed package or layer and checks the code against them. - Project Checks: Runs the project's lint, type-check, and test commands and requires failures to be fixed before proceeding. - Cross-Layer Review: Traces data flow (Storage → Service → API → UI), checks code reuse, import correctness, and same-layer consistency when changes span multiple layers. - Use Case: After implementing a new API endpoint that touches the service and UI layers, run this check to confirm tests pass, types flow correctly between layers, no duplicated constants were introduced, and the spec docs are updated with any lessons learned. ## Quick Start Run a quality check on my recent uncommitted changes and fix any violations you find.

Frequently Asked Questions about trellis-check

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check code quality before committing changes?▼

Run a structured review that inspects git diff output, executes the project's lint, type-check, and test commands, and verifies the changes against documented specs. Fix any failures and re-run the checks before committing.

How to verify data flow across multiple application layers?▼

Trace read flows from Storage through Service and API to UI, and write flows in reverse. Confirm types and schemas pass correctly between layers and that errors propagate properly to the caller.

When should cross-layer consistency checks be skipped?▼

Skip cross-layer checks when the change is confined to a single layer. They are only needed when modifications touch three or more layers, create new files, or modify shared constants and utilities.

Why update spec documentation after fixing a bug?▼

Documenting non-obvious bugs and discovered patterns in the spec prevents repeating the same mistake in future sessions. If a fix revealed something non-obvious, record it in the relevant spec file.

What should I check when creating new utility code?▼

Search the codebase for existing similar implementations before writing new ones. If two or more places define the same value, extract it into a shared constant, and verify all occurrences are updated after batch modifications.