GroupScribe
GroupScribe is an AI-powered assistant for LINE group chats that silently indexes work conversations, extracts tasks, decisions, and announcements into a structured format. It runs in the background without interrupting team communication.
Ficha actualizada 23 sept 2026, 14:38 · upstream 23 sept 2026, 14:17 · release sin release detectada
Qué es GroupScribe
Descripción editorial construida a partir del repositorio, su documentación y sus señales de mantenimiento.
GroupScribe transforms chaotic LINE group discussions into organized work records. Instead of letting information drift into the chat history, it quietly processes messages to identify what was said, who participated, and what actions were taken. The system extracts tasks with responsible assignees, calendar events from meeting notes, decisions made during calls, and documents like price quotes or specifications extracted from images and PDFs. All results are stored in a searchable index that employees can query directly through LINE.
- Extracting task assignments from informal group discussions into structured to-do lists with responsible assignees
- Converting meeting notes and call transcripts into calendar events and action items
- Organizing decisions made during calls or meetings into a searchable decision log
- Processing price quotes, specifications, and documents extracted from images or PDFs in LINE messages
- Creating announcements about upcoming events, deadlines, or policy changes based on group discussions
- Maintaining a searchable knowledge base of past conversations for reference and future queries
- Silent operation: the bot runs in the background without interrupting team communication or requiring manual intervention
- Complete transparency: all code is open source under AGPL-3.0, allowing teams to audit and modify the system themselves
- Self-hosted option available with full data sovereignty — messages are stored on your own infrastructure rather than a third-party platform
- Currently only supports LINE Messaging API; other platforms like Telegram or Discord require custom connectors
- AI extraction accuracy varies by input quality, and the system does not guarantee 100% precision in all cases
- Media files (images/PDFs) may take additional time to process due to OCR and analysis requirements
- The system only indexes messages from the last 30 days for new task extraction; older conversations remain searchable but cannot be re-extracted
- GPS-based attendance tracking is not fully reliable as it can be spoofed by browsers, so manual verification remains necessary
Capturas del proyecto
Imágenes reales detectadas en la documentación del repositorio.
Perfil de infraestructura
Estimación inicial derivada del repositorio y de los servicios detectados.
Despliegue detectado
Qué señales reales de instalación hemos encontrado.
GroupScribe requires a long-running always-on server to handle webhook processing and background AI inference. For self-hosting, use Docker with Tailscale Funnel for secure tunneling (e.g., `tailscale funnel --bg 3000`). The service must run in the same process as the webhook handler since it performs async LLM extraction that cannot be handled by a separate worker. Always-on infrastructure is required because processing time ranges from seconds to several minutes depending on message volume and AI model usage.
Stack y dependencias
Componentes inferidos a partir de manifests, configuración y documentación.
Puede sustituir a
Relaciones detectadas por el análisis del repositorio y su categoría.
Experiencia real de la comunidad
Ayuda a contrastar la estimación automática con instalaciones reales.
Todavía no hay datos comunitarios para este proyecto.