polym-lark-doc-ingest

Ingest Lark documents into a knowledge base with outline-first classification and desensitization.

8|Updated May 13, 2026
One-click install
npx skills add https://github.com/byteplus-sa/polym --skill polym-lark-doc-ingest
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: polym-lark-doc-ingest
Source: https://github.com/byteplus-sa/polym/tree/main/skills/polym-lark-doc-ingest
Command: npx skills add https://github.com/byteplus-sa/polym --skill polym-lark-doc-ingest

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It eliminates the manual process of tracking newly created or updated Lark documents and turning them into organized, searchable knowledge entries.

Core Features & Use Cases

  • Time-range discovery with deduping: Scans documents edited or created within a default period (or a user-specified range) and avoids reprocessing already-ingested items.
  • Outline-first classification to reduce cost: Fetches outlines first, classifies relevance, and only performs full fetch for qualifying documents.
  • Extraction + desensitization: Extracts TL;DR, customer/product/people, feature asks, decisions, competitive intel, and business progress, then applies desensitization rules before writing.
  • Knowledge-base writing (local + Lark): Saves raw snapshots and source/entity updates to the local wiki and updates the Lark wiki through the polym-sa-wiki write workflow.
  • Single-doc ingest mode: Ingests one specified document URL immediately without running discovery.

Quick Start

Run polym-lark-doc-ingest to ingest yesterday’s Lark documents into the knowledge base.

Frequently Asked Questions about polym-lark-doc-ingest

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

FAQPage Schema
How do I automatically sync Lark documents into a knowledge base?

To automatically sync Lark documents into a knowledge base, you can use a time-range discovery workflow that scans newly created or modified Lark docs, extracts structured insights, and writes them to local and Lark wiki storage.

How does outline-first document ingestion reduce processing costs?

Outline-first document ingestion reduces costs by fetching only the document outline first to classify relevance, then performing a full fetch and information extraction exclusively on qualifying documents.

Can I ingest a single specific Lark document URL into my wiki?

Yes, you can ingest a single specific Lark document URL immediately using the single-doc ingest mode, which bypasses the time-range discovery process and directly extracts structured insights.

How do I handle desensitization when extracting customer and competitive intel from Lark docs?

Desensitization is applied automatically during the extraction process, ensuring that extracted customer, product, people, and competitive intelligence is sanitized before writing to local wiki storage and Lark update queues.

How to avoid reprocessing already ingested Lark documents?

To avoid reprocessing already ingested Lark documents, the ingestion workflow applies deduplication rules that skip previously processed items during the time-range discovery phase.

What structured insights can I extract from Lark documents for knowledge organization?

You can extract TL;DR, customer/product/people entities, feature asks, decisions, competitive intel, and business progress from Lark documents to organize your knowledge base effectively.