AI Agents for Development#
Note
This section is a work in progress and may be updated frequently as the AI agent strategy and implementation is refined.
PyMAPDL uses structured AI agents to automate and improve development, testing, documentation, and code review. These agents are available in any LLM-enabled IDE (GitHub Copilot, Cursor, JetBrains AI, etc.) and are configured in the .github/agents/ directory.
Available Agents#
Developer: Implements features, fixes bugs, optimizes performance, and designs APIs.
Documentation Specialist: Reviews and improves documentation, validates docstrings, checks spelling and style.
Tester: Maintains test coverage, implements mocking strategies, and ensures test infrastructure quality.
Reviewer: Performs comprehensive PR reviews and enforces quality gates before merging.
Workflow#
The typical workflow is:
Developer implements code and features
Tester adds or updates tests, ensuring ≥90% coverage
Documentation Specialist reviews and improves documentation and docstrings
Reviewer checks code quality, documentation, and tests before merge
Agent Usage#
You can invoke agents by referencing their names in prompts or using slash commands
in supported IDEs. See .github/HOW_TO_USE_AGENTS.md for detailed instructions
and examples for each IDE.
Agent Guidelines#
Each agent follows a detailed specification in .github/agents/ and summarized in
AGENTS.md.Agents enforce project standards for code style, documentation, testing, and review.
See
AGENTS.mdand.github/HOW_TO_USE_AGENTS.mdfor full details and usage tips.
For more information, see:
.github/HOW_TO_USE_AGENTS.mdAGENTS.md