Agent skill

trellis-bootstrap

Initialize Trellis AI workflow system in a project. Use when user mentions "Trellis", "初始化 Trellis", "setup Trellis", "install Trellis", or wants to add AI-assisted development workflow to their project. This is a one-time setup skill.

View SKILL.md on GitHub Repository
Stars 5,124
Forks 287

Install this agent skill to your Project

npx add-skill https://github.com/mindfold-ai/Trellis/tree/main/.trellis/tasks/archive/2026-01/01-19-readme-redesign-taosu/bootstrap-skill

SKILL.md

Trellis Bootstrap

One-time setup for Trellis - the AI workflow system that helps your AI assistant remember project context across sessions.

What is Trellis?

Trellis provides:

  • Structure - Store coding guidelines in .trellis/structure/ that AI follows
  • Memory - Track session history in .trellis/agent-traces/
  • Automation - Slash commands like /start, /parallel, /finish-work

Prerequisites Check

Before starting, verify:

bash
# Check Node.js (required: v18+)
node --version

# Check npm
npm --version

If Node.js is not installed, ask the user to install it first.

Installation Steps

Step 1: Install Trellis CLI

bash
npm install -g @mindfoldhq/trellis@latest

Verify installation:

bash
trellis --version

Step 2: Get Developer Name

Ask the user for their name/username. This will be used for tracking their sessions.

Example question: "What name should I use for your developer profile? (e.g., your GitHub username)"

Step 3: Initialize in Project

Run in the project root:

bash
trellis init -u <developer-name>

This creates:

.trellis/
├── workflow.md                # Start here
├── structure/                 # Development guidelines
│   ├── frontend/
│   └── backend/
├── agent-traces/<name>/       # Your session history
└── scripts/                   # Automation scripts

.claude/
├── commands/                  # 13 slash commands
├── agents/                    # 6 agent definitions
└── hooks/                     # Automation hooks

.cursor/
└── commands/                  # 12 slash commands

AGENTS.md                      # AI reads this first

Step 4: Verify Setup

bash
# Check created files
ls -la .trellis/
ls -la .claude/commands/

Post-Setup Instructions

Tell the user:

  1. Start using Trellis - Run /start at the beginning of each session
  2. Add guidelines - Edit files in .trellis/structure/ to customize AI behavior
  3. Track progress - Run /record-agent-flow at the end of sessions

Quick Reference

Command When to Use
/start Beginning of every session
/parallel Complex features (multi-agent pipeline)
/before-frontend-dev Before frontend coding
/before-backend-dev Before backend coding
/finish-work Before committing
/record-agent-flow End of session

Success Criteria

Setup is complete when:

  • trellis --version shows version number
  • .trellis/ directory exists with workflow.md
  • .claude/commands/ contains slash command files
  • AGENTS.md exists in project root

After successful setup, this skill is no longer needed. The user should use /start to begin working with Trellis.

Note: This is a bootstrap skill. Once Trellis is initialized, use the built-in /start command instead of this skill.

Recommended Agent Skills

Expand your agent's capabilities with these related and highly-rated skills.

mindfold-ai/Trellis

create-command

Scaffolds a new skill file with proper naming conventions and structure. Analyzes requirements to determine skill type and generates appropriate content. Use when adding a new developer workflow skill, creating a custom skill, or extending the Trellis skill set.

5,124 287
Explore
mindfold-ai/Trellis

finish-work

Pre-commit quality checklist covering lint, typecheck, tests, code-spec sync, API changes, database migrations, cross-layer verification, and manual testing. Blocks commit if infra or cross-layer specs lack executable depth. Use when code is written and tested but not yet committed, before submitting changes, or as a final review before git commit.

5,124 287
Explore
mindfold-ai/Trellis

update-spec

Captures executable contracts and coding knowledge into .trellis/spec/ documents after implementation, debugging, or design decisions. Enforces code-spec depth for infra and cross-layer changes with mandatory sections for signatures, contracts, validation matrices, and test points. Use when a feature is implemented, a bug is fixed, a design decision is made, a new pattern is discovered, or cross-layer contracts change.

5,124 287
Explore
mindfold-ai/Trellis

check

Validates recently written code against project-specific development guidelines from .trellis/spec/. Identifies changed files via git diff, discovers applicable spec modules, runs lint and typecheck, and reports guideline violations. Use when code is written and needs quality verification, to catch context drift during long sessions, or before committing changes.

5,124 287
Explore
mindfold-ai/Trellis

check-cross-layer

Post-implementation verification across multiple code dimensions: cross-layer data flow, code reuse analysis, import path validation, and same-layer consistency checks. Identifies missed update sites, type mismatches, and duplicated constants. Use when changes span 3+ architectural layers, after modifying shared constants or configs, after batch file modifications, or when creating new utility functions.

5,124 287
Explore
mindfold-ai/Trellis

start

Initializes an AI development session by reading workflow guides, developer identity, git status, active tasks, and project guidelines from .trellis/. Classifies incoming tasks and routes to brainstorm, direct edit, or task workflow. Use when beginning a new coding session, resuming work, starting a new task, or re-establishing project context.

5,124 287
Explore

Didn't find tool you were looking for?

Be as detailed as possible for better results