v1.1.0: Conditional Browser Automation & Preflight Tools

[1.1.0] - 2026-09-04

Added

• Conditional browser automation with agent-browser as the preferred adapter and playwright-cli as fallback.
• Run-level authorization, source-import and module-finalization batch gates, bounded retries, and observable checkpoints.
• A deterministic Node/runtime and browser-tool preflight command with behavior tests and coverage thresholds.

Changed

• The 14-stage workflow now defaults to auto mode when a supported browser adapter is ready, while preserving guided mode and existing build-state compatibility.
• Course state now records automation mode, adapter, run identity, checkpoint, pending gate, and blocker.
• The beginner tutorial now starts with terminal, Node.js, skill, adapter, preflight, and first-sign-in completion criteria; fallback setup and troubleshooting remain behind targeted links.

v1.0.0: Initial Release

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog,
and this project adheres to Semantic Versioning.

[1.0.0] - 2026-09-03

Added

• Core Skill (SKILL.md):
  • Full 14-phase wizard-style methodology for transforming coarse course outlines into structured, citation-grounded curriculum in Google NotebookLM.
  • Strict distinction of three NotebookLM objects (Course Outline Source, Research Prompt, Candidate Source).
  • 3-Mode syllabus ingestion: Template Mode, Raw Text Ingestion Mode, and 3-Question Interview Mode.
  • Progress header tracking and session continuation mechanism (/resume).
  • Positive-steering behavioral constraints and verifiable completion criteria per step.
• Cross-Platform Agent Support (agents/openai.yaml):
  • Full compatibility with the open Agent Skills standard (agentskills.io, skills.sh).
  • Native integration with OpenAI Codex and ChatGPT App desktop/workspace environments.
• Reference Manuals (references/):
  • notebooklm-ui-guide.md: Detailed mapping of NotebookLM 3-panel UI layout (Sources, Chat, Studio), bilingual button tables, and step-by-step UI paths.
  • checklist.md: 14-stage verification checklist and key anti-patterns to avoid.
  • research-and-sources.md: Fast Research prompt engineering, 7-dimension source quality matrix, and Coverage analysis.
  • lesson-production.md: Single-lesson draft generation, zero-bloat in-notebook review loop, minor revisions, and final check rules.
  • module-completion.md: Module-level acceptance review, Quiz specifications, and Study Guide synthesis.
• Hands-on Tutorial (docs/first-course-tutorial.md):
  • Step-by-step onboarding walkthrough using a real-world Git collaboration course.
  • Beginner-friendly installation guide covering skills.sh, ChatGPT App, Claude Cowork, Google Antigravity, and Cursor.
  • AI-assisted installation prompt ("懶人救星") for zero-barrier setup.
• Automation Scripts & Templates (scripts/, templates/):
  • scripts/init-course.sh: Fast scaffolding for new course directories.
  • scripts/validate-state.sh: Automated inspection script for course build-state integrity.
  • templates/build-state.md: Single source of truth (SSOT) state ledger tracking research, coverage, drafts, reviews, and decisions.
  • templates/course-outline-template.md: Standard input format for course design.
• Real-World Examples (examples/):
  • examples/python-async-course/: Complete end-to-end course build demonstration covering syllabus, state ledger, and prompt dialogue history.
• Project Documentation:
  • Comprehensive bilingual documentation in Traditional Chinese (README.md) and English (README.en.md).