lattice

Explain Lattice fleet control and governance workflows for AI coding agents.

Updated Feb 19, 2026
One-click install
npx skills add https://github.com/plattegruber/docs --skill lattice
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: lattice
Source: https://github.com/plattegruber/docs/tree/main
Command: npx skills add https://github.com/plattegruber/docs --skill lattice

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The skill centralizes the Lattice control plane documentation so operators can quickly understand how to manage fleets of Sprites, govern Intents, and keep AI coding agents within safety boundaries.

Core Features & Use Cases

  • API reference coverage: Highlights fleet, sprite, intent, run, and health endpoints with authentication, payload, and error details.
  • Workflow guidance: Breaks down intents, approvals, safety classifications, runtime protocol events, and GitHub HITL steps for operators.
  • Use case: When you need to wake a Sprite, propose a task, approve a controlled intent, and observe the run, this skill maps each step to the right endpoint and policy.

Quick Start

Ask the skill to outline the Lattice endpoints, classifications, and steps needed to safely assign a task to a Sprite.

Frequently Asked Questions about lattice

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

FAQPage Schema
How do I manage a Lattice fleet and safely assign tasks to Sprites?

To manage a Lattice fleet, you map each operational step—waking a Sprite, proposing a task, and approving an intent—to the correct REST endpoint and safety policy. The workflow relies on fleet, sprite, intent, run, and health endpoints.

What is the Lattice runtime protocol for AI coding agents?

The Lattice runtime protocol defines event semantics that keep AI coding agents within safety boundaries. It processes runtime events, safety classifications, and approvals to govern agent actions during a task run.

How do I propose and approve Intents through the Lattice API?

You propose and approve Intents by sending request payloads to the documented Lattice intent endpoints. The workflow requires applying safety classifications and triggering GitHub HITL steps before a controlled intent is fully authorized.

Do I need prior API knowledge to govern Lattice safety workflows?

Yes, governing Lattice workflows requires familiarity with documented REST endpoints, request payloads, safety classifications, and runtime protocol event semantics to accurately guide AI coding agents and observe runs.

What endpoints are available in the Lattice API reference for fleet operators?

The Lattice API reference covers fleet, sprite, intent, run, and health endpoints. It provides operators with authentication details, request payload structures, and error specifics needed to orchestrate Sprites.

When should I apply safety classifications when orchestrating Sprites?

Safety classifications must be applied during the intent proposal phase before task execution. This ensures the Lattice control plane validates the proposed action against safety boundaries and runtime protocol policies.