Skip to content

AGENTS.md

Latest commit

 

History

History
126 lines (88 loc) · 4.2 KB

AGENTS.md

File metadata and controls

126 lines (88 loc) · 4.2 KB

AGENTS.md — AI Agent Guide for Microsoft Fabric CLI

This document provides guidance for AI agents assisting with the Microsoft Fabric CLI (fab) — a Python-based command-line interface for Microsoft Fabric.

About Microsoft Fabric CLI

The Fabric CLI is a file-system-inspired command-line interface that lets users explore, automate, and script their Microsoft Fabric environment. Key characteristics:

  • Installation: pip install ms-fabric-cli
  • Python versions: 3.10, 3.11, 3.12, 3.13
  • Platforms: Windows, macOS, Linux
  • Shells: PowerShell, Bash, Zsh, cmd
  • Modes: Interactive (REPL) and command-line (scripting)

Repository: https://github.com/microsoft/fabric-cli

AI Assets

All AI-related resources (skills, context, prompts, modes) are organized in the .ai-assets/ folder. See .ai-assets/README.md for a complete catalog of available resources.

Key Resources

Resource URL
AI Assets .ai-assets/README.md
Documentation https://microsoft.github.io/fabric-cli/
Command Reference https://microsoft.github.io/fabric-cli/commands/
Usage Examples https://microsoft.github.io/fabric-cli/examples/

For AI Agents Helping Contributors (Developing the CLI)

When assisting with code contributions to this repository (adding features, fixing bugs, writing tests), refer to the contributor context in .ai-assets/context/.

Forking and Setting Up

Guide users through the contribution workflow:

  1. Fork the repository: https://github.com/microsoft/fabric-cli/fork
  2. Clone locally: 
    git clone https://github.com/<username>/fabric-cli.git
cd fabric-cli
  3. Set up development environment (dev container recommended)
  4. Create a feature branch: 
    git checkout -b feature/my-change

Contribution Process

All PRs must follow this process:

  1. Find or create an issue — Search GitHub Issues first
  2. Look for help-wanted label — Required before starting work
  3. Comment on the issue — Describe your planned approach and wait for acknowledgment
  4. Link PR to issue — Start PR description with - Resolves #issue-number
  5. Add changie entry — Run changie new and select the appropriate change type

Code Standards to Enforce

When reviewing or generating code, ensure:

Requirement Details
Type hints All functions must have proper type annotations
Formatting Must pass black src/ tests/
Type checking Must pass mypy src/ tests/ --ignore-missing-imports
Naming snake_case for functions/variables, PascalCase for classes
Imports Grouped: stdlib → third-party → local
Copyright All new files need Microsoft copyright header
Docstrings Required for public functions

Error Handling Pattern

Always use structured error classes:

from fabric_cli.errors.common import CommonErrors
from fabric_cli import fab_constant

# Good
raise FabricCLIError(
    CommonErrors.invalid_path(path),
    fab_constant.ERROR_INVALID_INPUT
)

# Bad — hardcoded message
raise FabricCLIError("Invalid path provided")

Testing Requirements

All new functionality must include tests:

# Unit tests
python3 -m pytest tests/test_core tests/test_utils

# Integration tests (with VCR playback)
python3 -m pytest tests/test_commands --playback

Restricted Areas

Warn contributors that these areas require team involvement:

  • Authentication module — Security implications require Fabric CLI team review
  • Core infrastructure — Major architectural changes need team discussion

Quick Reference Links