MCP server

Your footage, queryable by Claude.

The Orcah Studio MCP server hands Claude, Claude Code, Cursor, and ChatGPT Codex 20+ tools over your indexed library, semantic search across three vector spaces, entity and metadata filters.

Works with Claude DesktopWorks with Claude CodeWorks with CursorWorks with ChatGPT Codex

Search

Three vector spaces. Picking the right one is most of the work.

text_searchWhat happens, who's in it, what's said — descriptions, transcripts, faces, objects, OCR

visual_searchWhat the frame looks like — CLIP text-to-image with sharpness and quality filters

audio_searchWhat it sounds like — music, ambience, sound, not words

Orient

Know what exists before you search. Never guess a story the footage can't support.

library_overviewMonths, people, locations, objects — what was actually shot, per year

entity_searchOne record per real person, place, sport, or device, with screen-time stats

index_statsIndex health — tells an empty index apart from a missed query

list_sourcesEvery indexed file — the vocabulary for scoped search

Read one video

From gist to full timeline, cheapest read first.

get_video_summary2–4 sentence summary plus chapter map

get_video_chaptersDeterministic chapter spans, always available

get_video_scenesThe whole timeline in capture order, with joined transcript

filename_searchResolve a half-remembered filename to its scenes

Filter by metadata

Exact ranges, no vectors — a range means a range.

camera_searchDrone / phone / action / professional families and recording modes

cinema_searchSony MXF cine data: ISO, aperture, focal length, colour, timecode

speaker_searchWho is speaking, narrowed by topic

face_emotion_searchWhose face is on screen, optionally by emotion

person_momentsOne person, on screen and speaking about a topic, in the same scene

telemetry_searchSpeed, motion, altitude, GPS bounding boxes

strava_searchHeart rate, distance, grade, sport from matched activities

Look and export

Judge visually, then land on the timeline.

contact_sheetUp to 48 thumbnails in one labelled grid — pick the best shot by looking

send_to_resolveAppend chosen clips to the open DaVinci Resolve timeline

get_resolve_timelineRead back the timeline name, fps, and clips

Typical flows

Documentary in an afternoon

library_overview(2026) → pick a narrative the footage supports → browse_scenes by month → contact_sheet → send_to_resolve

The exact line

text_search("she explains the pricing") → get_video_scenes(file) → full transcript, no paging

Best b-roll

visual_search(…, max_per_source=1, min_sharpness=100) → contact_sheet → pick a cell

Get Orcah Studio for Mac

Requires an Apple Silicon Mac, macOS 14.0 or later

Public beta — $249 lifetime license30-day refund, no questions asked