minutes-ingest

Extract structured facts from meeting transcripts into knowledge base profiles and logs.

1.4k|146|Updated Mar 18, 2026
One-click install
npx skills add https://github.com/silverstein/minutes --skill minutes-ingest
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: minutes-ingest
Source: https://github.com/silverstein/minutes/tree/main/tooling/skills/sources/minutes-ingest
Command: npx skills add https://github.com/silverstein/minutes --skill minutes-ingest

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Meetings produce decisions, action items, and context that are hard to surface later; minutes-ingest automates extracting those facts and updating a structured knowledge base so information is discoverable and actionable.

Core Features & Use Cases

  • Structured Extraction: Reads meeting YAML frontmatter and extracts decisions, action_items, entities, and intents with confidence and provenance.
  • Knowledge Updates: Appends timestamped log entries, updates person profiles, and maintains an index while deduplicating existing facts.
  • Adapters & Backfills: Supports PARA, Obsidian, and wiki adapters, and can process single meetings or backfill entire archives with a dry-run preview.
  • Safety & Validation: Honors confidence thresholds, skips low-confidence facts, records provenance, and recommends dry-run before writing.

Quick Start

Run minutes ingest --all --dry-run to preview extracted facts and ensure your knowledge adapter is configured.

Frequently Asked Questions about minutes-ingest

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I extract action items and decisions from meeting transcripts into Obsidian?

Meeting transcript extraction parses YAML frontmatter and text to pull decisions, action items, and entities directly into an Obsidian knowledge base. It appends timestamped log entries and updates person profiles while tracking provenance.

What is the best way to backfill an entire archive of meeting notes into a PARA knowledge base?

Backfilling a meeting archive processes historical files in bulk to populate a PARA knowledge base with structured facts. The system deduplicates existing entries and recommends running a dry-run preview to validate extracted data before writing.

Can I preview extracted meeting facts before committing them to my knowledge base?

Yes, you can run a dry-run preview to validate extracted meeting facts before writing. This safety mode honors configurable confidence thresholds and skips low-confidence facts, ensuring only accurate data populates your knowledge base.

How does meeting ingestion handle duplicate facts across multiple notes?

Meeting ingestion uses deduplication to prevent redundant facts across multiple notes. It cross-references new extracted data against existing knowledge base entries, appending only unique timestamped log entries and updating person profiles accordingly.

Do I need an LLM engine to process meeting files and update my knowledge base?

An LLM engine is optional for transcript-level extraction but not required for standard meeting files. The system reads YAML frontmatter natively, using configurable confidence thresholds and provenance tracking to safely update your index.