Deep requirements and design discovery workflow that keeps Trellis task-first PRD capture but adds stronger design confirmation, explicit technical design notes, and a clean handoff to task plan generation. Use when you want a richer brainstorming experience than $brainstorm, when trade-offs need deliberate review, or when you want Superpowers-style design validation inside Trellis tasks.

2026-05-038

check-backend

yyyzl/bilinote

Check if the code you just wrote follows the backend development guidelines.

Cross-Layer Check

Check if the code you just wrote follows the frontend development guidelines.

2026-05-038

name	create-command
description	Create New Skill

Create New Skill

Create a new Codex skill in .agents/skills/<skill-name>/SKILL.md based on user requirements.

Usage

$create-command <skill-name> <description>

Example:

$create-command review-pr Check PR code changes against project guidelines

Execution Steps

1. Parse Input

Extract from user input:

Skill name: Use kebab-case (e.g., review-pr)
Description: What the skill should accomplish

2. Analyze Requirements

Determine skill type based on description:

Initialization: Read docs, establish context
Pre-development: Read guidelines, check dependencies
Code check: Validate code quality and guideline compliance
Recording: Record progress, questions, structure changes
Generation: Generate docs or code templates

3. Generate Skill Content

Minimum SKILL.md structure:

---
name: <skill-name>
description: "<description>"
---

# <Skill Title>

<Instructions for when and how to use this skill>

4. Create Files

Create:

.agents/skills/<skill-name>/SKILL.md

5. Confirm Creation

Output result:

[OK] Created Skill: <skill-name>

File path:
- .agents/skills/<skill-name>/SKILL.md

Usage:
- Trigger directly with $<skill-name>
- Or open /skills and select it

Description:
<description>

Skill Content Guidelines

[OK] Good skill content

Clear and concise: Immediately understandable
Executable: AI can follow steps directly
Well-scoped: Clear boundaries of what to do and not do
Has output: Specifies expected output format (if needed)

[X] Avoid

Too vague: e.g., "optimize code"
Too complex: Single skill should not exceed 100 lines
Duplicate functionality: Check if similar skill exists first

Naming Conventions

Skill Type	Prefix	Example
Session Start	`start`	`start`
Pre-development	`before-`	`before-frontend-dev`
Check	`check-`	`check-frontend`
Record	`record-`	`record-session`
Generate	`generate-`	`generate-api-doc`
Update	`update-`	`update-changelog`
Other	Verb-first	`review-code`, `sync-data`