- Overview
- Architecture
- Installation
- Quick Start
- Configuration Reference
- Agents
- Directory Map
- Supporting Files
- Tests
- Contributing
- License
- Changelog
- Troubleshooting / FAQ
- Roadmap & Contact
This repository automates meeting documentation: transcript processing, agenda generation, and meeting minutes creation. It is designed for researchers, admins, and teams who want to streamline meeting records with minimal manual effort.
Data flow:
- Transcript ingestion →
scripts/transcript2json.py - Structuring →
meeting_utils.py(e.g.,parse_transcript()) - Minutes generation →
scripts/json2word.py
See doc/ for more details and templates.
pip install -r requirements.txt
# For DOCX export: install LibreOffice or Microsoft Word
# macOS: brew install --cask libreoffice# Convert transcript to JSON
python scripts/transcript2json.py sample.vtt > meeting.json
# Generate Word minutes from JSON
python scripts/json2word.py meeting.json
# Launch the GUI
python meeting_secretary_gui.pySee config.ini for all options. Example keys:
| Section | Key | Default | Description |
|---|---|---|---|
[general] |
language | en | Language for output |
[output] |
template | template_agenda | Template file for minutes |
[paths] |
output_dir | ./output | Where to save generated files |
You can override most options with environment variables (see README.md).
- Purpose: Main graphical user interface for the Meeting Secretary AI system.
- Functionality: Provides a user-friendly interface to interact with the meeting documentation tools, including transcript processing and minutes generation.
- Purpose: Utility functions for meeting data processing.
- Functionality: Contains helper functions used by other scripts for tasks such as parsing, formatting, and data manipulation.
- Purpose: Converts meeting minutes in JSON format to Microsoft Word documents.
- Functionality: Reads a JSON file (following the minutes schema) and generates a
.docxfile using a template.
- Purpose: Converts meeting transcripts to structured JSON format.
- Functionality: Processes raw transcript text and outputs a JSON file suitable for further processing or conversion to minutes.
- Purpose: Automates the process of generating meeting minutes from transcripts.
- Functionality: Shell script that chains together transcript processing and document generation steps.
scripts/– CLI tools for conversion and automationdoc/– Documentation, templates, and architecture diagrams__pycache__/– Python bytecode cache (ignored)requirements.txt– Python dependenciesconfig.ini– Main configuration filemeeting_secretary_gui.py– Main GUImeeting_utils.py– Utility functions
requirements.txt: Lists Python dependencies for the project.config.ini: Configuration file for customizing agent behavior.doc/: Documentation and templates for agenda and context generation.scripts/minutes_schema.JSON: JSON schema for meeting minutes structure.
Tests use pytest. To run:
pytestSee CONTRIBUTING.md for guidelines. Please:
- Use feature branches (
feature/your-feature) - Run
black .andflake8before PRs - Sign the CLA if prompted
SPDX-License-Identifier: MIT
See LICENSE for details.
See CHANGELOG.md for release notes and version history. Releases follow semantic versioning.
- Missing FFmpeg or LibreOffice: Install via
brew install ffmpeg libreoffice - Permission errors: Ensure scripts are executable:
chmod +x scripts/*.sh - Encoding errors: Use UTF-8 encoded input files
For more, see doc/FAQ.md or open an issue.
Vision: Streamline and automate meeting documentation for research and professional teams. Planned: web interface, more export formats, and AI-powered summarization.
See GitHub Issues for the roadmap.