Installation notes, configuration, and practical tool use.
Agent buildingAgent projects, workflows, tool connections, and implementation notes.
Agent evaluationTest cases, evaluation methods, reliability, and quality questions.
Field notesTool trials, lessons learned, and notes from the field.
A focused workspace for turning requirements into reviewable test assets.
Latest notes
Engineering guide
A practical guide to building a maintainable, testable, and packageable Python project with a src layout, pyproject.toml, pytest, Ruff, CLI commands, Makefile, and GitHub Actions.
Agent building · Reference project
From natural language to the real macOS Calendar: an engineering case study of a Mac-first personal agent.
Agent evaluation
A practical framework for testing an agent across deterministic tests, evaluation, observability, integration, E2E, and regression.
Agent evaluation · Case study
How testing for trading, fund transfers, and a Calendar Agent can evolve into measurable, evidence-based AI-assisted QA.