data-ingest

Ingest raw text and export files into an Obsidian wiki.

4|Updated May 9, 2026

Search Tags:#safety #text processing #wiki #knowledge distillation #ingest #obsiidan #chat exports

Ready to run

npx skills add https://github.com/itsmarsss/memento --skill data-ingest-itsmarsss

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: data-ingest
Source: https://github.com/itsmarsss/memento/tree/main/obsidian-wiki/.skills/data-ingest
Command: npx skills add https://github.com/itsmarsss/memento --skill data-ingest-itsmarsss

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill solves the problem of converting unstructured or unfamiliar raw text sources into organized Obsidian wiki pages you can actually use.

Core Features & Use Cases

Universal text ingestion for chat exports, conversation logs, transcripts, emails, bookmarks, CSV-like dumps, and other raw text collections.
Format-aware parsing that identifies common structures (JSON/JSONL, Markdown, plaintext, CSV/TSV, HTML-like exports, chat log patterns).
Knowledge distillation and page clustering that extracts topics, decisions, facts, procedures, and entities, then deduplicates across sources.
Safety and trust boundary controls that treat source content as untrusted data and never executes embedded “instructions.”
Manifest and activity updates to track what was ingested, which pages were created/updated, and to keep the wiki index current.

Quick Start

Ask the skill to ingest this data by pointing it to the file or folder path you want imported into the Obsidian wiki.

Frequently Asked Questions about data-ingest

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

FAQPage Schema

How do I convert chat exports and raw text logs into an Obsidian wiki?▼

To convert chat exports and raw text logs into an Obsidian wiki, the Skill ingests source files, detects formats like JSON or Markdown, and distills content into structured pages by extracting topics, decisions, and entities.

What is the best way to extract topics and decisions from unstructured text exports?▼

Extracting topics and decisions from unstructured text exports requires clustering and deduplicating information across sources. This Skill identifies facts, procedures, and entities, generating organized wiki pages while treating all input content as untrusted data.

Can I import CSV dumps and email archives directly into a knowledge base?▼

Yes, you can import CSV dumps and email archives directly into a knowledge base. The Skill utilizes format-aware parsing to identify structures in CSV, TSV, plaintext, and HTML-like exports, transforming them into wiki pages.

Does the text ingestion process protect against embedded instructions in chat logs?▼

Yes, text ingestion protects against embedded instructions in chat logs by enforcing strict safety boundary controls. It treats all source content as untrusted data, preventing any embedded instructions from executing or exfiltrating information.

What file formats are supported for ingesting raw text into an Obsidian vault?▼

Supported file formats for ingesting raw text include JSON, JSONL, Markdown, plaintext, CSV, TSV, and HTML-like exports. The Skill detects formats automatically and processes mixed folders containing images via vision capabilities.

How do I prevent duplicate wiki pages when importing multiple conversation logs?▼

To prevent duplicate wiki pages when importing multiple conversation logs, the Skill uses manifest and idempotency checks alongside clustering and deduplication. It tracks ingested sources and updates the wiki index to avoid redundant content.