Mento – AI‑powered triage tool for open‑source maintainers

Описание и применение

Mento – инструмент для автоматической приоритизации и суммирования GitHub‑issues, использующий LLM и Sanity для обучения на прошлых решениях.

Задача

Maintainers of large legacy codebases struggle to triage the high volume of daily GitHub issues, leading to missed priorities and wasted effort.

Результат

Mento reduces manual triage effort, provides consistent urgency scoring, and learns from maintainer actions to improve priority suggestions over time.

Рабочий процесс

• 1. Connect Mento to a GitHub repository (no GitHub login required, use an LLM key).
• 2. Mento fetches open issues via the GitHub API.
• 3. For each issue, an LLM generates a concise summary, urgency score (1‑10), key points, and suggested actions.
• 4. The triage pipeline stores data in Sanity, learning from past maintainer decisions.
• 5. Mento presents issue cards with summaries, urgency, and priority suggestions (P0–P4).
• 6. Maintainers review and act on the suggested priorities, updating status in Mento.
• 7. Each triage decision feeds back into Sanity, refining future suggestions.

Доказательства

Mento fetches your GitHub issues generates plain English summaries with urgency scores and runs an intelligent triage pipeline that learns from past decisions to suggest P0–P4 priorities.

Источник

The more issues you triage the smarter it gets.

Источник