Tool reference

The core private-preview workflow is search, inspect the matching passage, then retrieve the full source with citations.

`search_content`

Hybrid keyword + semantic search with evidence snippets.

Param	Use
`query`	Search topics, names, or tickers. Empty to browse.
`source`	Show, publication, or newsletter name.
`mode`	`hybrid` (default) · `keyword` · `semantic`.
filters	`speakers, people, companies, tickers, topics, shows, authors, publications, content_types, source_types, languages` — alias-resolved.
window	`published_after / published_before` (ISO dates).
`sort`	`relevance` (recency-blended) or `recent`.

“Latest from X”: query="", source="X", sort="recent". Results carry best_snippets with speaker and timestamp offsets. Review these before fetching full text.

`get_content`

Record by ID: full reading text, attribution metadata, summary, and canonical source_url. Search results contain bounded timestamped evidence passages; fetch the full record only when needed.

`request_ingestion`

Queue a public episode, YouTube, RSS, or newsletter URL. Returns an existing content_id or new job_id. Text takes minutes; audio is transcribed on-device overnight, searchable the next morning. Content joins the shared corpus.

`get_ingestion_status`

Poll a job: queued → working → completed for text; needs_local → local_fetching → prefetched → completed for overnight audio.

`list_watchlist`

Monitored shows, newsletters, and people, with counts and latest coverage. Checked every 15 minutes.

Recipes

Earnings-season sweep

search_content("NVIDIA earnings", published_after=…) → inspect the evidence passages → get_content for context.

Track a person

search_content(query="AI spending", speakers=["Brad Gerstner"]) → compare his timestamped passages across sources and dates. Speaker filters include only current speaker-scoped transcripts; review the returned attribution_coverage note.

Ingest and wait

request_ingestion(url) → for queued audio, advise waiting until morning. For text, poll get_ingestion_status every ~20s for up to 5 minutes.

Brief me on a theme

search_content("AI capex sustainability", sort="relevance") → compare evidence passages across shows and newsletters → fetch the strongest sources for complete context and citations.

Machine-readable: server card · llms.txt · setup guide