intel-intake

To turn external articles and links into actionable intel, ingest the content to cross-reference named entities in your workspace. The system reads entities.json and events.jsonl to surface precise connections and saves structured outputs to _hq/intel.

Entity-aware content ingestion processes external text, links, or videos to anchor information to named entities like people and projects. It cross-references your workspace events.jsonl and entities.json to replace generic categories with precise, deduplicated intel.

Yes, you can ingest video content to map it to dormant customers and open commitments. The ingestion process loads BUSINESS_CONTEXT.md, aliases.json, and KNOWLEDGE_BASE.md to ensure the resulting intel aligns with your workspace routing.

The intel ingestion workflow requires BUSINESS_CONTEXT.md, entities.json, aliases.json, and KNOWLEDGE_BASE.md. Loading these files ensures proper alignment, deduplication, and appropriate workflow routing before generating results.

The best way to avoid duplicating existing knowledge base entries is to load KNOWLEDGE_BASE.md alongside aliases.json during ingestion. This ensures deduplication and appropriate workflow routing before cross-referencing results into entities.json.

Structured intel outputs are saved to the _hq/intel directory. The system stores results with cross-references into entities.json for relevant threads, ensuring the ingested external content connects to existing workspace data.