Skip to content

Latest commit

 

History

History
34 lines (23 loc) · 1.31 KB

AGENTS.md

File metadata and controls

34 lines (23 loc) · 1.31 KB

Repository Agent Guidance

This repository provides two portable Agent Skills for working with Modules 0-4. Load the relevant SKILL.md when the user's request matches it.

Before applying either skill, read .agents/course-context.md. It defines the course scope, important files, ExecuTorch API taxonomy, repository structure, and the code that should receive the most attention.

Explain Mode

For requests to learn, explain, review, trace, compare, or test understanding, read and follow:

.agents/skills/explain-mode/SKILL.md

Keep the interaction conversational. Inspect the published course content and corresponding repository scripts or application code, explain the ExecuTorch APIs, and ask focused questions that let the learner check and deepen their understanding.

Build Mode

For requests to build, adapt, debug, extend, or prototype an ExecuTorch model pipeline or application, read and follow:

.agents/skills/build-mode/SKILL.md

Preserve the course's clear separation between model preparation, lowering, validation, and runtime integration. Explain major ExecuTorch API boundaries as the implementation progresses. Before adapting a course example into a new prototype, verify the design against current official ExecuTorch, PyTorch, platform, and backend documentation.

Treat Modules 0-4 as the course.