trellis-meta

Guides customization of local Trellis workflow, hooks, agents, and platform files.

1|Updated Jan 22, 2026
One-click install
npx skills add https://github.com/MaplumeX/Milesto --skill trellis-meta-maplumex
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: trellis-meta
Source: https://github.com/MaplumeX/Milesto/tree/main/.agents/skills/trellis-meta
Command: npx skills add https://github.com/MaplumeX/Milesto --skill trellis-meta-maplumex

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? After running trellis init, AI assistants often lack a map of the generated .trellis/ directory and platform integration files, leading to edits in the wrong place (such as node_modules or upstream source) or broken workflow semantics. This Skill gives the AI a complete model of the local Trellis architecture and the correct entry point for every customization request. ## Core Features & Use Cases - Architecture Reference: Explains the three-layer local model (workflow, persistence, platform integration) covering .trellis/workflow.md, config.yaml, tasks, specs, workspace journals, and runtime state. - Platform File Map: Documents where hooks, settings, agents, skills, commands, and prompts live for Claude Code, Cursor, Codex, Copilot, Kiro, Gemini, Windsurf, and other platforms. - Customization Playbooks: Step-by-step guides for changing workflow phases, task lifecycle hooks, context injection, agent behavior, spec structure, and adding project-local skills. - Use Case: A user says "the implement agent never reads my backend spec." The AI reads the context-loading reference, checks the task's implement.jsonl, and fixes the JSONL manifest or the inject-subagent-context hook instead of guessing. ## Quick Start Ask the AI to use the trellis-meta skill to change the no-task workflow rule in .trellis/workflow.md so task creation is skipped for one-reply questions.

Frequently Asked Questions about trellis-meta

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

FAQPage Schema
How do I customize the Trellis workflow in my project?▼

Edit .trellis/workflow.md, which is the local workflow source of truth defining phases, skill routing, and workflow-state prompt blocks. After editing, check that platform skills, commands, and agents still describe the same flow.

How do I change what context AI agents read before implementation?▼

Configure the task's implement.jsonl and check.jsonl files, which list spec and research files agents must read first. On hook-push platforms edit the inject-subagent-context hook; on agent-pull platforms edit the agent file's read steps.

Which AI platforms does Trellis support for hooks and agents?▼

Trellis generates platform files for Claude Code, Cursor, OpenCode, Codex, Kiro, Gemini CLI, Qoder, CodeBuddy, GitHub Copilot, Factory Droid, and Pi Agent, plus workflow-based platforms like Kilo, Antigravity, and Windsurf. Actual directories depend on which trellis init flags were run.

Why does task.py current report no active task in my shell?▼

Active task state is stored per session in .trellis/.runtime/sessions/ and resolved from platform events or environment variables like TRELLIS_CONTEXT_ID. If the shell lacks session identity, check the platform's shell session bridge hook rather than creating a global pointer file.

Should I edit node_modules or the Trellis source to change behavior?▼

No. Default to editing generated files inside the user project: .trellis/workflow.md, config.yaml, spec, scripts, and platform directories. Only touch upstream source when explicitly contributing a PR, forking, or changing npm publish contents.

Where should team coding conventions live in a Trellis project?▼

Put long-term engineering conventions in .trellis/spec/ with an index.md per layer, or create a project-local skill for team-specific flows. Do not store project-private rules in the public trellis-meta skill, since updates could overwrite them.