Frapea

Measured conclusion: Deep-tested. L2 outcome: tools_ok.

A video editor in your browser that your AI drives: cuts, captions, multicam, motion titles, export.

Add to your agent

Copy this into your MCP client config:

{
  "mcpServers": {
    "Frapea": {
      "url": "https://app.frapea.com/mcp",
      "transport": "streamable-http"
    }
  }
}

Measured tools (67)

• connect

  Pairs you with the user's Frapea browser tab. TWO WAYS ROUND. (1) If they already have Frapea open: ask them for the pairing code (the Connect MCP button) and call this with it. (2) If they do not, or you are not sure: call this with NO arguments and you get back a link — send them the link, then call this again with the code it contains every few seconds until it returns a token (they have to accept it in their browser first). Either way you end up with a session token to pass on every later call. Codes are single-use and expire in ten minutes; a token lasts until the user disconnects.

  Parameters:

  • code — string — The 8-character code — from the user's screen, or from the link this tool gave you. Omit it to get a link to send them.

  Hints: read-only no · destructive no · idempotent no · open-world no

• get_projects read-only

  Lists the projects Frapea knows on this machine: id, name, lastOpenedAt, and for a project stored in a folder on disk its `folder` (the directory's name — files you write into its `media/` subfolder join the pool by themselves). Call first to find the projectId every other tool needs. A project the browser has no permission for will surface PERMISSION_REQUIRED when used — relay the returned instruction to the user, then retry. ALSO RETURNS `providers` — Pexels (stock video/photos), Freesound (music, ambience, SFX) and ElevenLabs (voiceover), each 'connected' | 'checking' | 'missing' | 'invalid'. These are the user's OWN API keys and they cannot be set from here. When any is not usable, `providersAdvice` is present: a short note for the user saying which and where they connect them (only the user can add a key). When everything is connected it is absent. A call that fails with PERMISSION_REQUIRED carries its own instruction for the user.

  Parameters:

  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• get_timeline read-only

  Reads a project's timeline structure: timelines (id, name, fps, resolution) with tracks and clips. THE USER CAN COPY A CLIP'S ID from the editor's right-click menu, so a bare UUID in their message names exactly one clip — call this with NO timelineId to search every timeline and match `clip.id`, rather than guessing from a name. Clips carry id, name, mediaRef, kind, startFrame, endFrame (exclusive), trimStartFrame and linkGroupId (linked A/V pairs share it), plus `look` (opacity, blend, transform, crop, keyframes) whenever any of it differs from the defaults — an untouched clip has no `look` key. Omit timelineId to get every timeline; frames are in that timeline's fps.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • timelineId — string — Optional single timeline to read.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• get_media read-only

  Lists the project's media library: mediaRef (the id clips use), name, kind (video/audio/image), durationSeconds, resolution, and sceneSamples — timestamps (seconds) where the visual indexer detected distinct moments. sceneSamples is a ready-made shot map: inspect or cut at those times first. Each asset also carries visualIndex and transcript objects: {phase: 'ready'|'pending'|'queued'|'running'|'failed'|'no-audio'|'disabled', error, attempts, sealed}. Analysis starts by itself and retries a failure a few times before sealing it. If a phase is 'failed', read `error` and decide: call retry_analysis to run just that one analysis again, or leave it. If 'disabled', call request_ai_models. Video and audio assets also carry beats: {phase: 'ready'|'pending'|'failed'|'none', bpm, confidence}. These come from an ONSET detector (spectral flux): every energy rise — drum hits, plucks, syllables — not bars or downbeats. bpm is the tempo of those onsets, or null when there is no steady pulse (confidence 0–1 below the detector's 0.3, or too few onsets) — speech has onsets and no bpm. 'none' = no audio to analyze. Pass mediaRef to get ONE asset with beats.onsetsSec (source seconds, the first 2000; truncated + onsetCount say when there are more) — map them through the clip's trim and speed yourself.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • mediaRef — string — Optional: return only this asset, as {asset}, with its onset times. Omit to list every asset (no onset times).

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• get_transcript read-only

  Returns the word-level speech transcript of one media asset: language and words [{text, startSec, endSec}]. Only assets the user has transcribed have one — NOT_FOUND means ask the user to run Transcribe on the clip in Frapea. Use the timestamps to align cuts with what is being said.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • mediaRef (required) — string — Asset reference from get_media.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• create_motion_composition

  Creates a MOTION COMPOSITION — an animated graphic (infographic, title, promo end-card, chart…) written as a React TSX component. Read the 'motion' skill (read_skill {topic:'motion'}) for the exact API surface, available fonts/icons and golden examples before writing one. files maps file names to TSX source; the entry must `export default` the component. schema declares the editable props (type string|number|color|boolean|select|mediaRef + default) — they become Inspector controls the user can tweak, so parameterize colors, copy and media slots. The composition is compiled and test-rendered BEFORE saving; on failure the error returns verbatim — fix the code and retry. Returns the mediaRef to place with add_clips. Creative guidance is in read_skill {topic:'motion'}. With UI-like parts (buttons, pills, progress bars, cards), write them in their own file with a partsBoard prop (default false) and look at the board first: preview:true, previewProps:{partsBoard:true} returns one large frame with the reply.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • name (required) — string — Display name, e.g. 'Pink promo end card'.
  • files (required) — object — File name → TSX source. Relative imports between them work.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • fps — number — Composition fps (default 30).
  • entry — string — Entry file (default composition.tsx); must export default.
  • width — number — Composition width px (default 1920).
  • height — number — Composition height px (default 1080).
  • schema — object — Editable props: {key: {type, default, label?, min?, max?, options?, multiline?}}. type mediaRef makes a media-pool slot.
  • preview — boolean — Return one large frame (≤1560 px) with the reply. Media is not loaded in a preview: media slots draw empty and the reply says how many.
  • previewFrame — number — Frame to preview (default 0).
  • previewProps — object — Merged over the schema defaults for the preview picture only, e.g. {partsBoard:true}. Nothing saved changes.
  • durationInFrames — number — Length in frames (default 150).

  Hints: read-only no · destructive no · idempotent no · open-world no

• update_motion_composition destructive

  Updates a motion composition's TSX files and/or manifest (name, duration, schema…). Same compile-and-test gate as create; files given here MERGE over the existing set. Every clip playing the composition re-renders. preview:true returns one large frame of the result with the reply.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • compositionId (required) — string — From create/list_motion_compositions.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • fps — number
  • name — string
  • files — object — File name → TSX source (merged).
  • width — number
  • height — number
  • schema — object
  • preview — boolean — Return one large frame (≤1560 px) with the reply. Media is not loaded in a preview: media slots draw empty and the reply says how many.
  • previewFrame — number — Frame to preview (default 0).
  • previewProps — object — Merged over the schema defaults for the preview picture only, e.g. {partsBoard:true}. Nothing saved changes.
  • durationInFrames — number

  Hints: read-only no · destructive yes · idempotent yes · open-world no

• get_motion_composition read-only

  Returns a motion composition's manifest and full TSX sources.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • compositionId (required) — string — Composition id.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• list_motion_compositions read-only

  Lists the project's motion compositions: id, mediaRef (for add_clips), name, size, fps, duration and the editable props schema.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• create_timeline

  Creates a new timeline in the project and returns its id. Set fps, width and height for the target format (e.g. 1080×1920 @30 for a vertical short). The new timeline starts with one video and one audio track; spare tracks appear automatically as clips land.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • name (required) — string — Display name, e.g. 'TikTok short'.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • fps — number — Optional frames per second (default 30).
  • width — number — Optional width in px (default 1920).
  • height — number — Optional height in px (default 1080).

  Hints: read-only no · destructive no · idempotent no · open-world no

• add_clips destructive

  Places media segments on a timeline. Each clip: mediaRef (from get_media, 'timeline:<id>' to NEST another timeline as a clip, 'solid:#rrggbb' , 'motion:<id>' for a MOTION composition (create_motion_composition) for a generated solid-colour matte, 'adjustment:' for an adjustment clip — no picture of its own, its effects/masks/opacity grade everything below it — or 'text:' for a title, restyled via set_clip_properties' `text`), atFrame (timeline position), sourceInSec/sourceOutSec (the segment of the source to use — align these with transcript words or sceneSamples), and optional trackId (from get_timeline) to target a specific same-kind track — without it clips land on V1/A1 with overwrite semantics. A spare empty track always exists above the highest used one, so V2/A2 is available before anything sits on it. Video sources with audio get a linked audio clip automatically (on the matching A track: V2 ⇒ A2). Returns the created clip ids in order. Frames are in the target timeline's fps.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Target timeline id from get_timeline.
  • clips (required) — array — [{mediaRef, atFrame, sourceInSec, sourceOutSec, trackId?}] — segments in playback order; trackId targets a specific track (default V1/A1).
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive yes · idempotent no · open-world no

• split_clips

  Blades clips at a timeline frame. Splits every listed clip (and its linked partner) into two independent clips at atFrame. Frames are in the target timeline's fps. Use with get_timeline to find clip ids first.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline containing the clips.
  • clipIds (required) — array — Clip ids to split.
  • atFrame (required) — number — Timeline frame to cut at (must lie inside each clip).
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive no · idempotent no · open-world no

• remove_clips destructive

  Deletes clips (lift — leaves a gap). Linked partners are NOT removed automatically; pass both ids of a pair to delete it whole. To close the gap too, use ripple_delete_ranges instead.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline containing the clips.
  • clipIds (required) — array — Clip ids to delete.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive yes · idempotent yes · open-world no

• ripple_delete_ranges destructive

  Extracts [fromFrame, toFrame) on one track and shifts everything after it left to close the gap (sync-locked tracks follow). The go-to tool for tightening a cut or removing a bad take.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Target timeline.
  • trackId (required) — string — Track id from get_timeline.
  • fromFrame (required) — number — Range start (inclusive).
  • toFrame (required) — number — Range end (exclusive).
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive yes · idempotent no · open-world no

• set_clip_properties destructive

  Adjusts a clip. Only the provided fields change. Mix: volume (0–2, 1 = unity), fadeInFrames / fadeOutFrames (gain or opacity ramps at the edges). Look: opacity (0–1), blend (normal | multiply | screen | overlay | darken | lighten | color-dodge | color-burn | hard-light | soft-light | difference | exclusion | add), transform, crop and speed. transform.fit picks how the picture meets the frame before scaling: contain (letterbox, default) | cover (fill + crop overflow) | fill (stretch, ignores aspect). Identity = contain fit; position is offset from centre in timeline pixels (y down), scale multiplies the fitted size, rotationDeg is clockwise about the anchor (a fraction of the cropped picture, 0.5/0.5 = centre). Crop takes fractions off each edge of the source. Retiming: `speed` (constant rate/direction) and `speedKeyframes` (source-anchored rate curve; reflows length). Shape masks: `masks` (full-list replace). `disabled` parks a clip (it stays but neither renders nor sounds — DaVinci's D). `text` restyles text clips (mediaRef 'text:'). `motion` customizes MOTION clips (mediaRef 'motion:<id>') per clip.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline containing the clip.
  • clipId (required) — string — Clip to modify.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • crop — object — Optional partial crop, fractions 0–1 of the SOURCE off each edge (the inspector shows source pixels). Crop is a window: it cuts pixels away without moving or re-fitting the picture; edges may meet (the picture vanishes).
  • text — object — Optional TEXT clip restyle (clips with mediaRef 'text:'; create them with add_clips). Partial patch: content ( breaks lines), fontFamily (Inter | DM Sans | Space Grotesk | Bebas Neue | Playfair Display | Caveat), sizeFraction (of frame height), bold, italic, color (#rrggbb), align (left|center|right), outlineColor (#rrggbb or null), outlineWidth, backgroundFill (#rrggbb[aa] or null), box {x,y,w,h in frame fractions, centre-based}, animation. wholeCaptionGroup: true restyles every caption sharing the clip's caption group in one edit (content stays per-clip).
  • blend — string — Optional blend mode (see description).
  • masks — array — Optional shape masks, FULL-LIST replace (empty array removes all). Masks UNION (Resolve's additive Power Windows); `invert` flips one mask's coverage. Geometry in fractions of the SOURCE picture, centre 0.5/0.5 — masks ride the clip's transform. Each: {shape: rect|ellipse, center{x,y}, size{x,y}, rotationDeg, feather (fraction of source width), opacity 0–1, invert}.
  • speed — object — Optional retiming (Resolve's Speed Change). rate 1 = 100 % (0.01–50); direction forward | reverse | freeze; freezeFrame = held source frame (offset from trim-in) while frozen; pitchCorrection keeps the audio's pitch. Keyframes stay glued to the pictures (they are source-anchored). Combine with speedRipple: true to rescale the clip so the same source range plays at the new rate and shift everything downstream.
  • bypass — object — Optional section switches, true = section OFF (its values are kept but not applied) — the inspector's per-section toggles.
  • motion — object — Optional MOTION clip customization (clips with mediaRef 'motion:<id>'). props: partial patch over the composition's schema-declared props ({key: value}; a key set to null resets it to the composition default). get_motion_composition lists the schema; per-clip values override its defaults.
  • volume — number — Optional new volume (0–2).
  • denoise — object — AI voice denoise (dry/wet). Setting it (or enabling the section) queues the cached background render for the clip's source file.
  • opacity — number — Optional static opacity 0–1.
  • disabled — boolean — Optional. True disables the clip (and its A/V link partners): it stays on the timeline but neither renders nor sounds. False re-enables. The multicam angle switch is built on this.
  • transform — object — Optional partial transform; provided keys replace, others stay.
  • cropRetain — boolean — Optional. True = Resolve's 'Retain Image Position': the crop window stays fixed in the frame while transforms move the picture underneath it.
  • motionBlur — object — Optional SHUTTER MOTION BLUR from the clip's own movement: each frame is the average of `samples` sub-frame renders spread over the open shutter, each at the position/scale/rotation (keyframes, dynamic zoom, tracking) the clip has at that instant — so a fast move streaks along its path and a still clip stays sharp (and costs nothing extra). {enabled: true} (or just true) turns it on — 6 samples at 180°, the film look, unless the clip already has its own; {enabled: false} (or false) turns it off; a partial setting changes only what it names. Costs one extra draw per sample on frames where the clip moves. Movement INSIDE the picture (footage, a motion composition's own animation) is not blurred — only the clip's transform. Not to be confused with the 'motionBlur' EFFECT (manage_effects): that is a fixed directional blur filter, the same on every frame; use this setting for blur that comes from the clip actually moving.
  • solidColor — string — Optional, solid-colour matte clips only: recolour the matte (#rrggbb). Mattes are created by add_clips/insert_clips with mediaRef 'solid:#rrggbb'.
  • dynamicZoom — object — Optional Ken Burns move (Resolve's Dynamic Zoom): animates the framing from `start` to `end` over the clip's WHOLE duration (retimes when the clip is trimmed). Rects: centre x/y as picture fractions (0.5/0.5 = middle) and scale (1 = whole picture, 0.5 = half → 2× zoom). Setting framing or ease switches it on.
  • speedRipple — boolean — Ripple Timeline for this speed edit (see `speed`).
  • cropSoftness — number — Optional crop-edge feather, fraction of the source width 0–0.5 (the inspector shows source pixels). Keyframable as 'cropSoftness'.
  • fadeInFrames — number — Optional fade-in length in frames.
  • fadeOutFrames — number — Optional fade-out length in frames.
  • stabilization — object — Optional stabilization knobs (Resolve's Stabilization). The expensive motion tracking is a per-media analysis — run it with the stabilizer; these values post-process its cached path instantly. mode perspective | similarity | translation; cameraLock kills ALL motion (ignores ratio/smooth); zoom scales up to hide blanking; croppingRatio 0.25–1 (1 allows no correction); smooth 0.25–1; strength -1–1 (negative inverts).
  • lensDistortion — number — Optional Lens Correction Distortion, -1–1. Positive straightens wide-angle barrel distortion, negative adds it. 0 = none.
  • speedKeyframes — array — Optional source-anchored speed keyframes (forward-direction clips only): [{atSourceFrame, rate}], rate 1 = 100 %. Rate is CONSTANT between points (held outside). Present, it replaces `speed.rate` AND REFLOWS the clip's timeline length to keep the same source window (slower ⇒ longer); the A/V link group retimes together. EMPTY array clears the curve (back to the constant rate).

  Hints: read-only no · destructive yes · idempotent yes · open-world no

• generate_captions

  Generates animated captions from a spoken clip's transcript (the same generator as Frapea's UI): transcript words group into caption text clips over the clip's span, word-timed so the animation follows speech. One undo, and they share a captionGroupId — restyle them all via set_clip_properties with text.wholeCaptionGroup. Needs a transcript (get_transcript tells you). Returns the created clip ids. PLACEMENT: they JOIN the existing caption track whenever they fit in its gaps, and only open a new one when they would overlap (a second language, a second layer). So captioning four narration segments in four calls leaves ONE caption track, not four — no tidying needed afterwards for this. Call manage_tracks {action:'tidy'} at the end of the build for the empty tracks left by everything else. They join the LOWEST caption track, so a picture layer you add ABOVE it after an earlier pass will cover them — captions render behind b-roll rather than looking wrong, which is the hardest kind of mistake to notice. Move the whole lane with manage_tracks {action:'merge'} if you want the captions on top.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline containing the spoken clip.
  • clipId (required) — string — The spoken clip to caption.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • preset — string — Style preset: clean (default) | boldPop | karaoke | highlight | minimal | editorial.
  • position — string — bottom (default) | top.
  • textCase — string — original (default) | upper | lower.
  • censorProfanity — boolean — Star out profane words.
  • maxWordsPerCaption — number — Words per caption before a new one starts (default 4).

  Hints: read-only no · destructive no · idempotent no · open-world no

• manage_effects destructive

  Edits a clip's effect stack (ordered, per-effect enable). action: 'add' (type; returns the new effect's id) | 'remove' (effectId) | 'move' (effectId + toIndex — the order IS the render order) | 'toggle' (effectId + enabled) | 'set' (effectId + params patch, clamped to the effect's ranges) | 'curve' (effectId + curve + points). Types: 'colorBasic' (Lumetri Basic Correction-compatible: temperature/tint -100..100, exposure -5..5 stops, contrast/highlights/shadows/whites/blacks -100..100, saturation 0..300, vibrance -100..100), 'colorWheels' (Resolve lift/gamma/gain/offset — {group}Master/{group}R/G/B; lift/gamma 0 neutral, gain 1, offset printer points 25 neutral), 'lut' (.cube from the project library — set `resource`), 'curves' (tone curves), 'hueCurves' (hue-vs-hue/sat/lum). Also 'motionBlur' — NOT the shutter blur: a DIRECTIONAL blur filter (angle 0–360, distance 0–100) that smears every frame the same way whether the clip moves or not (a streak look, a speed effect on footage). For blur that follows the clip's own keyframed movement, use set_clip_properties motionBlur instead.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Target timeline.
  • clipId (required) — string — Target clip.
  • action (required) — string — add | remove | move | toggle | set | curve
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • type — string — Effect type for 'add'.
  • curve — string — For 'curve': which curve to replace. 'curves' effect: master|r|g|b (tone, identity = [0,0,1,1]); 'hueCurves': hueHue|hueSat|hueLum (x = hue with red at 0, y 0.5 = untouched).
  • params — object — For 'set': partial param patch, e.g. { exposure: 1.5 }.
  • points — array — For 'curve': flat control points [x0,y0,x1,y1,…] in the unit square.
  • enabled — boolean — For 'toggle'.
  • toIndex — number — Target position for 'move' (0 = applied first).
  • effectId — string — Effect instance id (from a previous add or get_timeline).
  • resource — string — For 'set' on resource effects ('lut'): the LUT's library name (a .cube imported into the project). Empty string clears it.

  Hints: read-only no · destructive yes · idempotent no · open-world no

• manage_transitions destructive

  Transitions on a track's cuts OR lone clip edges (a free head/tail takes a single-sided fade from/to nothing; its alignment is forced). action: 'add' (trackId + atFrame = the cut where one clip ends and the next starts, + type; optional durationFrames (default 30, clamped to the clips' source handles), alignment centered|start|end, params) | 'remove' (transitionId) | 'update' (transitionId + any of type/durationFrames/alignment/params). Types: crossDissolve, dipToBlack, dipToWhite, wipe (params.angle deg, params.softness 0-100), slide (params.angle), push (params.angle). Both clips need source material past the cut — a transition that cannot fit is refused; edits that break the cut remove it. Returns {ok, transitionId?}.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Target timeline.
  • trackId (required) — string — Track owning the cut (get_timeline).
  • action (required) — string — add | remove | update
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • type — string — Transition type.
  • params — object — Type params patch (e.g. { angle: 90 }).
  • atFrame — number — The cut frame for 'add'.
  • alignment — string — centered | start | end
  • transitionId — string — Existing transition id.
  • durationFrames — number

  Hints: read-only no · destructive yes · idempotent no · open-world no

• manage_tracks destructive

  Track operations on a timeline. action: 'mute' | 'unmute' | 'solo' | 'unsolo' (audio only; any solo silences non-solo tracks) | 'hide' | 'show' | 'rename' (needs name) | 'duck' | 'unduck' (audio only: dip this track under speech detected on the other audio tracks — transcript word timings when available, energy VAD otherwise) | 'merge' (needs intoTrackId: moves every clip of trackId onto another track of the same kind; refuses the WHOLE move if any clip would overlap, or if the source carries transitions) | 'delete' (an EMPTY track; refuses one that holds clips — deleting those is remove_clips, by name) | 'tidy' (NO trackId: drops the empty leftover tracks and renumbers). Track ids come from get_timeline. TIDYING UP, AND WHY ORDER MATTERS. Track counts only ratchet up while you build: a spare appears above whatever you fill, and the video and audio counts are kept EQUAL. That last rule is the one that surprises people — five video tracks force five audio tracks, however empty. So: 1. 'merge' the tracks that hold clips but need not be separate (captions from several passes, b-roll that never overlaps). This is what actually lowers the count. 2. 'tidy' once at the end. It settles both sides on one spare above the taller one and reports removedTracks. 'delete' on a single empty track usually reports removedTracks: 0 — the invariant puts it straight back to keep the counts equal. That is not a failure, it is why 'tidy' exists. Neither ever touches a clip, and both keep an empty track you renamed, muted, hid, soloed, unlocked or set to duck.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Target timeline.
  • action (required) — string — mute | unmute | solo | unsolo | hide | show | rename | duck | unduck | merge | delete | tidy
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • name — string — New name (rename only).
  • trackId — string — Track to modify. Required for every action except 'tidy'.
  • intoTrackId — string — Destination track (merge only) — same kind as trackId.

  Hints: read-only no · destructive yes · idempotent no · open-world no

• cut_to_angle

  MULTICAM angle switch. The angles are ordinary video tracks stacked over each other (sync them beforehand — clips of the same take on V1..Vn); this razors the whole angle stack at `atFrame` and keeps only `chosenTrackId` live from there to the next cut — the other angles' clips are DISABLED, not deleted, so every switch stays re-decidable (cut again with a different angle to change your mind). Linked audio follows the video by default (audioFollowsVideo: false keeps audio untouched — the usual choice when one good mic carries the sound). angleTrackIds defaults to every video track with a clip under the frame. Refused when the chosen track has no clip at the frame.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Target timeline.
  • atFrame (required) — number — The switch point (timeline frame).
  • chosenTrackId (required) — string — The angle to show from atFrame on.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • angleTrackIds — array — Optional explicit angle set (video track ids).
  • audioFollowsVideo — boolean — Default true: linked audio toggles with its angle.

  Hints: read-only no · destructive no · idempotent no · open-world no

• move_clips destructive

  Moves clips to new positions (linked partners follow automatically). Each move: clipId, toStartFrame, optional toTrackId (same-kind track). Overwrite semantics — landing on another clip trims it like a drag-drop.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Target timeline.
  • moves (required) — array — [{clipId, toStartFrame, toTrackId?}]
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive yes · idempotent yes · open-world no

• set_active_timeline

  Sets which timeline the project opens on (and what the user sees if the editor is open), opening its tab if it was closed. Purely a convenience — every tool here takes an explicit timelineId regardless.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline to activate.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive no · idempotent yes · open-world no

• set_project_settings destructive

  Updates a timeline's fps / width / height (clips keep their timing; content re-fits) and/or renames it. Only provided fields change.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline to update.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • fps — number — Optional new frames per second.
  • name — string — Optional new display name.
  • width — number — Optional new width in px.
  • height — number — Optional new height in px.

  Hints: read-only no · destructive yes · idempotent yes · open-world no

• search_media read-only

  Semantic search over the project's media. Visual scope matches what is IN the picture (describe a shot: 'aerial city at night'); spoken scope matches transcribed speech (exact words, ranked first, with a snippet). Hits: {mediaRef, timeSec (best moment), fromSec, toSec, score, scope}. Requires the user to have downloaded the picture analyzer — otherwise PERMISSION_REQUIRED tells you what to relay.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • query (required) — string — What to look for — plain language.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• remove_words destructive

  Text-based editing: cuts the given transcript words OUT of a clip as ripple deletes (everything after each cut shifts left). wordIndices are positions in get_transcript's words array for the clip's media. Contiguous indices merge into single cuts. The go-to tool for removing filler words or tightening an interview.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline containing the clip.
  • clipId (required) — string — The clip whose speech is being edited.
  • wordIndices (required) — array — Indices into get_transcript words, e.g. [4,5,6,12].
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive yes · idempotent no · open-world no

• remove_silence destructive

  Cuts silent gaps inside a clip: any pause between transcribed words longer than minGapSec becomes a ripple delete (a small margin is kept so speech never clips). Typical minGapSec: 0.6 tight, 1.0 balanced, 1.5 loose.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline containing the clip.
  • clipId (required) — string — The clip to tighten.
  • minGapSec (required) — number — Minimum pause length to cut (seconds).
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive yes · idempotent no · open-world no

• undo destructive

  Reverts the most recent edit THIS agent session made to the project (one step per call; repeat to go further). Cannot revert other tabs' or the user's edits.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive yes · idempotent no · open-world no

• inspect_media read-only

  Looks INSIDE a source asset. Returns ONE contact sheet — every requested moment as a labelled cell in a single grid image — plus a per-cell fingerprint. Pass atSec (max 12 timestamps, seconds); combine with get_media's sceneSamples for an instant storyboard of the whole file. Shows the RAW source; use inspect_timeline for the composed cut. START WITH pixels:false WHEN YOU ARE ONLY CHECKING FOR EMPTINESS. The fingerprint answers 'is this black / flat / did it decode' for no image cost — nonBlankRatio near 0 is an empty or black frame, colorRange near 0 is a featureless one. Ask for pixels when you must judge WHAT is in the frame: subject, framing, legibility.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • mediaRef (required) — string — Asset from get_media.
  • atSec (required) — array — Timestamps in seconds, max 12.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • pixels — boolean — False = fingerprints only, no image (default true).
  • columns — number — Grid columns (default 3, or one row if ≤3).
  • cellHeight — number — Target px height per cell (default 300). Shrunk automatically if the sheet would exceed the response budget — the reply says so when that happens.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• inspect_timeline read-only

  WATCHES the cut. Composites every video layer at each requested frame and returns ONE contact sheet — the frames as labelled cells of a single grid image — plus, per cell, the visibleClipIds that contributed pixels and a fingerprint. Pass frames (timeline frame numbers, max 12). Use it to check an edit: this is the only tool that shows what actually PLAYS, as opposed to what get_timeline says is arranged. SAMPLE THE CUT POINTS AND THE MIDDLE OF EACH SECTION, not only frame 0. An entrance animation that never resolves, a graphic clipped by its own box, an offline clip rendering black — all look perfect in get_timeline and are obvious here. USE pixels:false FOR A CHEAP SWEEP. visibleClipIds plus nonBlankRatio confirms 'the right clip is in every slot and none of them is black' across a dozen positions with no image cost; spend pixels on the few frames where you must judge typography, framing or timing.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline to render.
  • frames (required) — array — Timeline frame numbers, max 12.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • pixels — boolean — False = structure + fingerprints only, no image (default true).
  • columns — number — Grid columns (default 3, or one row if ≤3).
  • cellHeight — number — Target px height per cell (default 300). Shrunk automatically if the sheet would exceed the response budget — the reply says so when that happens.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• get_problems read-only

  What has gone WRONG in the user's editor — renders that failed or stopped, footage that would not decode, chunks that could not be prepared for playback, exports that failed — each with WHERE (the timeline and frames, the source and its own frames, the footage and frame the engine could not decode) and the reason in the engine's own words. You also hear about new problems for free: every other tool result carries a `problems` list of what changed since your last call. Call this for the whole history, or after a `moreProblems` count. A problem that is open on a timeline means that timeline does NOT play or export correctly there, whatever inspect_timeline showed: fix the named layer (e.g. update_motion_composition) to fix it. `state` is open | fixed | superseded (the thing it was about was edited and not yet re-rendered — NOT proof it works) | dismissed.

  Parameters:

  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • detail — boolean — Include the engine's raw diagnostic text per problem (default false — it can be long).
  • projectId — string — Project id from get_projects. Omit for every project in this browser.
  • includeResolved — boolean — Also list problems that are no longer open (default false).

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• get_bug_report

  Prepare a BUG REPORT for the user: it opens in Frapea (Problems → Create bug report) for them to read and save. The report itself is NOT returned to you — it would leave the user's machine before they have seen it. You get a summary: open problems, media counts, and how many of the user's own words remain in composition code. It contains versions, problems with engine details, the project's structure with names and texts replaced, provider files by public id, the user's own files only as encoding fingerprints, and motion code. timelineIds limits it to those timelines and what they nest.

  Parameters:

  • projectId (required) — string — Project id from get_projects (must be open in the editor).
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • timelineIds — array — Only these timelines and the timelines they nest (default: all).

  Hints: read-only no · destructive no · idempotent yes · open-world no

• verify_timeline read-only

  Renders a timeline the way an export does and reports what fails. inspect_timeline shows a few frames through the preview path, which quietly substitutes a nearby picture where a frame cannot be rendered; this renders every motion segment the timeline reads (through every nested timeline) exactly, then composites every frame with every layer (footage decoded), and fails where an export would. Returns {jobId}; poll get_job_status every ~5 s. When done, `result` is {verdict: ok | problems | incomplete, checked, notChecked, segments, chunks, problems: [...same shape as get_problems, with `related` warnings that often name the cause...], stale?, note}. `incomplete` is NOT a pass: something could not be checked. `stale: true` means the timeline was edited while it ran — verify again. Sound is not checked. Needs the project open in the editor tab you are paired with (error PROJECT_NOT_OPEN otherwise).

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline to verify (nested timelines inside it are verified too).
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • cancel — boolean — Stop the verification of this timeline that is running.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• insert_clips

  Ripple-inserts media segments: everything at/after atFrame shifts right to make room (clips under the cut split). Same clip shape as add_clips. Use to splice a shot INTO an existing sequence without overwriting.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Target timeline.
  • atFrame (required) — number — Insertion point (timeline frame).
  • clips (required) — array — [{mediaRef, sourceInSec, sourceOutSec}]
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive no · idempotent no · open-world no

• organize_media destructive

  Pool file operations. action: 'rename' renames the real file on disk (needs newName; clips referencing the old name go offline). 'delete' UNLINKS the file from the media pool — the real file on disk is NOT touched; clips using it go offline until it is added again.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • mediaRef (required) — string — Asset from get_media.
  • action (required) — string — rename | delete
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • newName — string — New file name incl. extension (rename).

  Hints: read-only no · destructive yes · idempotent yes · open-world no

• read_skill read-only

  Returns Frapea's editing documentation for a topic. With no topic: the editing conventions, how the tools fit together and the current feature surface. MOTION topics: 'motion' is the API surface (imports, fonts, props schema, the parts-board-first golden example) and routes to the rest — read before create_motion_composition; 'motion-examples' (two finished pieces that differ in look); 'motion-craft' (anticipation, overshoot, stagger, the tells of machine-made motion — a companion to 'motion'); 'motion-text' (size floors, fitting copy that cannot clip, per-word animation, annotations); 'motion-effects' (grain, duotone, light leaks, blurs, glow); 'motion-transitions' (TransitionSeries, presentations, cut overlays); 'motion-sound' (where sound lives, the built-in SFX pack, placement and levels); 'motion-review' (the same loop as 'review', from a motion job). The CRAFT guides teach how to actually cut: 'core' (rhythm, b-roll, audio hierarchy, restraint — read first), then the format you are making — 'shorts' | 'longform' | 'podcast' | 'tutorial' | 'montage'. Also 'voiceover' (narration in segments, not one block), 'verify' (how to check a cut before export) and 'review' (rounds of critique against rendered frames, for a piece with an audience).

  Parameters:

  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • topic — string — Optional. Motion: motion | motion-examples | motion-craft | motion-text | motion-effects | motion-transitions | motion-sound | motion-review. Craft: core | shorts | longform | podcast | tutorial | montage | voiceover | verify | review.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• await_user_action read-only

  Waits (up to ~45 s per call) for the user to resolve a pending Frapea prompt — e.g. the folder-permission dialog another tool just opened. The dialog waits for the user in the Frapea tab; the tool that opened it returned what they need to click. Returns {status: 'granted' | 'denied' | 'pending'}: 'pending' means they have not acted yet (call again to keep waiting), 'denied' that they declined.

  Parameters:

  • actionId (required) — string — The actionId a previous tool returned.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• open_project destructive

  Navigates the user's Frapea tab to a project so they can WATCH the edit live. Not required for editing — every tool works headlessly; use this for show-your-work moments. This reloads the Frapea tab, so an export or render running there stops.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive yes · idempotent yes · open-world no

• close_project destructive

  Navigates the user's Frapea tab back to the project manager and releases this session's headless runtime for the project. When the tab shows that project it reloads, so an export or render running there stops.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive yes · idempotent yes · open-world no

• new_project

  Creates a project. By default it is created in the browser's own storage, which needs no permission and no click: the project id comes back immediately and you can start working. The user can turn it into a real folder on their disk whenever they want (Cmd/Ctrl+S, or File → Save to a Folder) without losing anything. Pass storage:'folder' only when the user asked for a folder up front — that needs a browser security gesture, so it opens a dialog in Frapea with a 'Choose folder' button, returns an actionId, and you must tell the user IN YOUR REPLY to click it, then call await_user_action; on 'granted' the project appears in get_projects.

  Parameters:

  • name (required) — string — Project name, e.g. from the user's brief.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • storage — string — Where the project lives. 'browser' (default) needs no user action. 'folder' asks the user to pick a directory on their disk.

  Hints: read-only no · destructive no · idempotent no · open-world no

• connect_media_folder

  Asks the user to attach a folder of THEIR OWN source footage to the project — opens a dialog in Frapea with a 'Connect folder' button. Only a human can do this (browser security gesture): tell the user IN YOUR REPLY to click it, then call await_user_action with the returned actionId; on 'granted', get_media will list the folder's files. Use when get_media comes back empty and the user has footage somewhere on their disk. NEVER call it for files YOU made or fetched. The project folder's `media/` folder is scanned: any media file placed there joins the pool by itself (within a second, or when the user returns to the tab), with no dialog. `reason` is shown in the dialog word for word — say which footage you need and what you will do with it, so the user knows what to pick.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • reason (required) — string — One or two sentences for the dialog, addressed to the user: which folder to pick and why — e.g. 'Pick the folder with your trip clips from Lisbon, so I can cut them into the reel you asked for.'
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive no · idempotent no · open-world no

• trim_clips destructive

  Adjusts one edge of a clip (linked A/V partners follow). edge: 'start' trims/extends the in-point, 'end' the out-point. deltaFrames > 0 moves the edge right, < 0 left — e.g. edge 'end', delta -12 shortens the clip by 12 frames. Extending is limited by the source material.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline containing the clip.
  • clipId (required) — string — Clip to trim.
  • edge (required) — string — start | end
  • deltaFrames (required) — number — Signed frame delta for that edge.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive yes · idempotent no · open-world no

• get_analysis_status read-only

  How far the project's media analysis has got. Analysis runs BY ITSELF whenever media is connected. Returns percent (0-100) — 100 means nothing is OUTSTANDING, not that everything succeeded: `failed` lists items whose analysis failed (retry_analysis re-runs one) — done/total/pending item counts, runningItem {mediaRef, kind, progress}, etaSeconds (null until something has finished), blockedOnModels, and models {visual, speech}. Poll this while you wait; when blockedOnModels is above zero, call request_ai_models.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• request_ai_models

  Asks the user to enable Frapea's on-device AI analyzers (picture search, speech). Opens a dialog in every Frapea tab and returns an actionId — tell the user IN YOUR REPLY to confirm it, then call await_user_action and poll get_analysis_status until models are enabled. Call this when get_media reports 'needs-models'.

  Parameters:

  • projectId (required) — string — Project the work belongs to.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive no · idempotent no · open-world yes

• start_visual_indexing

  Indexes the project's pictures for search_media. Usually not needed: indexing is automatic and this only nudges the queue to re-check now. When the user switched automatic indexing off, this asks for every video and image to be indexed. Returns the same payload as get_analysis_status.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive no · idempotent yes · open-world no

• start_transcription

  Transcribes one asset even if automatic transcription is switched off. Speech transcripts otherwise appear on their own. Returns the project's analysis status; poll get_analysis_status, then read get_transcript.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • mediaRef (required) — string — Asset from get_media.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive no · idempotent yes · open-world no

• get_job_status read-only

  Status of a background job started by a job-returning tool (generate_voiceover, verify_timeline, measure_loudness): {state: queued|running|done|failed|cancelled, progress: 0-1, detail, ...result fields such as mediaRef when done; verify_timeline, measure_loudness puts its report in `result`}. Poll every few seconds while state is 'queued' or 'running' (queued = waiting for the analysis master tab, which loads the models once for the whole browser). For analysis started by start_transcription or start_visual_indexing, poll get_analysis_status instead.

  Parameters:

  • jobId (required) — string — Id returned by the start_* tool.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• list_voiceover_models read-only

  The text-to-speech providers Frapea can generate voiceovers with. Returns providers[]: the FREE in-browser models straight from the TTS registry ({id, label, languages, voices: [{id, label, language, gender, grade}], modelAssets, state: {phase: needs-download|downloading|ready|error}}), plus the PREMIUM cloud provider flagged premium: true ({id: 'elevenlabs', label, premium, voices: [{id, label}], models: [{id, label, creditsPerCharacter, supportsAudioTags, note}]}, voices/models empty until a key is stored). supportsAudioTags tells you whether that model understands inline audio tags like [calm], [laughs], [whispers] as delivery directions (only the v3 family does); on every other model such tags are READ ALOUD, so send those models plain text — each model's `note` restates its rule. Also returns a top-level status object `elevenlabs`: {provider: 'elevenlabs', hasApiKey, credits: {used, limit, remaining} | null, consent: 'ask' | 'always'} — READ IT BEFORE choosing the premium provider, so you know whether a key exists, what the account can afford and whether a consent dialog will be involved. A free provider whose phase is 'needs-download' cannot generate until the user enables it in the Generators panel (Voiceover) — there is no tool to force the download; ask the user.

  Parameters:

  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• generate_voiceover

  Generates an AI voiceover from a script — on-device with the free models, or in the cloud with ElevenLabs (the user's own key and credits). The WAV lands in the project's media library as a normal audio asset (with waveform and analysis), the script is stored as asset metadata, and a script-corrected word-level transcript is written so captions work immediately. Narration is best generated before the picture is cut: a script does not tell you how long it takes to SAY, so a narrated edit assembled first is built on guessed cut points that the real audio then breaks — every section, graphic and transition has to be retimed. Generate the narration, read its word timings with get_transcript, then lay picture against those frames. Segments read better than one block (e.g. hook / body / close, or one per section), each laid on the timeline separately. A single long generation comes out FLAT and evenly paced, everything pressed together at one energy; segmenting gives each part its own intent, puts natural air at the joins, and lets you redo one section instead of all of it. Write for the ear: short sentences, one idea each, LINE BREAKS between them — the strongest signal the model has to actually stop. See read_skill {topic:'voiceover'}. Async: returns {jobId, fileName} — fileName is the audio file's final name in the project's media/ folder, decided up front (WAV for local providers, MP3 for ElevenLabs). The job runs on the analysis MASTER tab and its request is persisted on disk: it survives tab closes and browser restarts, and get_job_status answers from disk even in a fresh session. Poll until done; the finished job also carries the asset's mediaRef. Voices/languages come from list_voiceover_models; a local model must be downloaded first (phase 'ready'), otherwise this returns PERMISSION_REQUIRED. PREMIUM (provider 'elevenlabs' — SPENDS THE USER'S CREDITS): check list_voiceover_models' `elevenlabs` status object first. Structured refusals you must handle: NO_API_KEY (no key stored — ask the USER to add their key in the Voiceover generator's ElevenLabs onboarding; never ask for, or pass, the key yourself), INSUFFICIENT_CREDITS (carries needed vs remaining — shorten the script or ask the user), CONSENT_REQUIRED (paid features are set to 'ask': a consent dialog with the cost is now open in Frapea and the error carries an actionId for the user to confirm; call await_user_action with that actionId, and on 'granted' retry this call with consentActionId set to it; 'denied' means drop it). AUDIO TAGS (ElevenLabs): inline bracket tags like [calm], [excited], [whispers] direct the delivery ONLY on models whose list_voiceover_models entry says supportsAudioTags: true (the v3 family). Every other model reads them ALOUD — so only write tags when the chosen model supports them. As a safety net, tags sent to a non-supporting model are stripped before synthesis and the success message says so.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • text (required) — string — The script to speak (multi-line ok).
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • speed — number — Speaking rate multiplier (default 1). Local providers accept 0.5–2; ElevenLabs 0.7–1.2.
  • voice — string — Voice id from the provider's manifest (default: its first voice for the chosen language).
  • modelId — string — ElevenLabs only: TTS model id from list_voiceover_models (default 'eleven_v3' — expressive, reads [calm]-style audio tags as directions; use 'eleven_multilingual_v2' only when the user asks for v2; turbo/flash cost half). Check the model's supportsAudioTags before writing [calm]-style tags into the text — non-v3 models would read them aloud.
  • fileName — string — Optional base name for the audio file written into media/.
  • language — string — Language tag from the provider's manifest (e.g. 'en-us'). Ignored for ElevenLabs — its voices are multilingual.
  • provider — string — TTS provider id from list_voiceover_models (default 'kokoro'; 'elevenlabs' = premium, needs the user's key + consent).
  • voiceSettings — object — ElevenLabs only: voice knobs, each optional. stability, similarityBoost and style are 0–1; speakerBoost is a boolean.
  • consentActionId — string — The actionId from a CONSENT_REQUIRED refusal, after await_user_action returned 'granted'. Good for one job.

  Hints: read-only no · destructive no · idempotent no · open-world yes

• search_stock_media read-only

  Searches the Pexels stock library (photos or videos) with the user's connected API key. Returns items with id, dimensions, duration (videos), author, thumbnail url, available video qualities, and a `downloaded` flag telling whether the asset is already in the project's media pool. If no key is connected it errors with PERMISSION_REQUIRED, whose message says where the user connects one (Library → Generators → Stock Library). Results are subject to the Pexels license; keep the author attribution when asked about provenance. Results match the uploader's keywords, not the picture — a search for 'earth from space' also returns the Moon and artists' renders; `thumbnail` shows the actual subject. See read_skill {topic:'verify'}.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • kind (required) — string — photos | videos
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • page — number — 1-based page, 40 items per page.
  • size — string — Optional: large | medium | small
  • color — string — Optional, photos only: a Pexels color name.
  • query — string — Search phrase; empty = curated/popular feed.
  • orientation — string — Optional: landscape | portrait | square

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• download_stock_media

  Downloads one Pexels asset into the project's media pool as a virtual file at the Library root (name `pexels-<id>.jpg|.mp4`). Asynchronous: the first call accepts the download and returns its state; CALL AGAIN with the same arguments to poll — state becomes 'done' with the asset's mediaRef, ready for add_clips. maxHeight picks the best video rendition not exceeding it (default: the best available). ASK FOR THE WHOLE SHOT LIST AT ONCE — three files stream at a time and the rest wait their turn, so a batch is queued for you, in the order you asked. State 'queued' is normal and needs no action but polling. ONCE DOWNLOADED, CHECK THE MOMENT YOU MEAN TO USE — not just the first frame. A clip that opens on your subject can cut away, change subject or hold a logo exactly where you planned to trim. Read the scene samples from get_media, or render frames with inspect_media, across the range you intend to cut. See read_skill {topic:'verify'}.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • kind (required) — string — photo | video
  • pexelsId (required) — number — The item id from search_stock_media.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • maxHeight — number — Videos only: cap the rendition height (e.g. 1080).

  Hints: read-only no · destructive no · idempotent yes · open-world yes

• search_sounds read-only

  Searches the Freesound library (SFX, ambience, foley) with the user's connected API key. Every item carries what you need to JUDGE the sound without hearing it: the uploader's full description, tags, duration, community rating and download count, format/samplerate, and — when analyzed — AudioCommons descriptors (acAnalysis: ac_loudness LUFS, ac_dynamic_range, ac_tempo, ac_tonality, ac_single_event, plus 0-100 timbre scales like ac_brightness, ac_warmth, ac_hardness, ac_depth, ac_roughness, ac_boominess, ac_sharpness, ac_reverb). A `downloaded` flag says the sound is already in the pool. Licenses: CC0 needs nothing, BY needs attribution, BY-NC excludes commercial use — prefer cc0/by via the license arg when the project may be commercial. If no key is connected it errors with PERMISSION_REQUIRED, whose message says where the user connects one (Library → Generators → Sound Library).

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • query (required) — string — Search phrase (e.g. 'door slam', 'jungle ambience').
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • page — number — 1-based page, 40 items per page.
  • sort — string — Optional: score (default) | rating_desc | downloads_desc | created_desc
  • license — string — Optional: cc0 | by | by-nc
  • maxDurationSeconds — number — Optional upper duration bound.
  • minDurationSeconds — number — Optional lower duration bound.

  Hints: read-only yes · destructive no · idempotent yes · open-world yes

• download_sound

  Downloads one Freesound sound (its HQ MP3 preview, ~128 kbps — plenty for SFX/ambience) into the project's media pool as a virtual file at the Library root (name `freesound-<id>.mp3`). Asynchronous: the first call accepts the download and returns its state; CALL AGAIN with the same arguments to poll — state becomes 'done' with the asset's mediaRef, ready for add_clips onto an audio track. Three files stream at a time and the rest wait their turn, shared with stock video, so state 'queued' is normal and needs no action but polling.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • soundId (required) — number — The sound id from search_sounds.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive no · idempotent yes · open-world yes

• add_sfx

  Frapea's BUILT-IN sound-effects pack — synthesized, licence-free, no API key, no download wait. Covers common UI and motion sounds; search_sounds is for what the pack does not (ambience, foley, a specific real-world object). Call with no `ids` to get the catalogue: each sound carries what it sounds like AND when to use it. Call with `ids` to copy them into the project's media pool as `sfx-<id>.wav` at the Library root; each result gives a mediaRef ready for add_clips on an audio track. Adding a sound already in the project is a no-op that returns its existing mediaRef.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • id — string — Convenience: a single sound id.
  • ids — array — Sound ids to add (e.g. ['whip','thump','ding']). Omit to list the pack.

  Hints: read-only no · destructive no · idempotent yes · open-world no

• manage_queue destructive

  Full control of the background work queue. action 'list' returns every queued/running item across all projects (analyses, proxies, beats) plus the visible task rows (downloads, voiceover jobs, failures) with a cancellable flag. action 'cancel' stops ONE kind of work for ONE asset — queued or already running (a running visual index stops at the next sample and keeps its checkpoint). The asset then shows an incomplete-analysis badge; retry_analysis or the tile's retry button runs it again. kind: index | transcribe | stabilize | denoise | proxy | beats.

  Parameters:

  • action (required) — string — list | cancel
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • kind — string — cancel only: index | transcribe | stabilize | denoise | proxy | beats
  • mediaRef — string — cancel only: asset from get_media.
  • projectId — string — cancel only: project id from get_projects.

  Hints: read-only no · destructive yes · idempotent no · open-world no

• retry_analysis destructive

  Runs ONE analysis of ONE asset again — picture indexing or speech transcription — clearing whatever failure was recorded for it. If the asset already has a finished transcript or index, that result is deleted and rebuilt. Use when get_media shows phase 'failed' and the error looks worth another go, or when you want a fresh result for a file that changed meaning. Returns the project's analysis status.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • mediaRef (required) — string — Asset from get_media.
  • kind (required) — string — index | transcribe
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive yes · idempotent yes · open-world no

• set_keyframes destructive

  Animates one clip property. property: opacity | volume | positionX | positionY | scaleX | scaleY | rotationDeg | pitch | yaw | anchorX | anchorY | cropLeft | cropTop | cropRight | cropBottom | cropSoftness. Keyframes are stamped in SOURCE frames (frames into the media, i.e. timelineFrame - startFrame + trimStartFrame), so they survive moves and trims. action 'set' (default) upserts the given keyframes; 'remove' deletes at the given frames; 'clear' removes the property's animation (or ALL animation when property is omitted). easing is how the value LEAVES a keyframe: linear | hold | smooth.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline containing the clip.
  • clipId (required) — string — Clip to animate.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • param — string — The effect param key to animate (e.g. 'exposure').
  • action — string — set | remove | clear (default set).
  • effectId — string — Animate an EFFECT param instead of a clip property: the effect instance id (from manage_effects add / get_timeline). Use with `param`; `property` is then ignored.
  • property — string — Animatable property (see description).
  • keyframes — array — For 'set': keyframes to add or replace.
  • atSourceFrames — array — For 'remove': source frames whose keyframes to delete.

  Hints: read-only no · destructive yes · idempotent yes · open-world no

• apply_layout destructive

  Arranges clips into a screen layout by writing ordinary transform + crop values (first clipId → first cell). layout: full | side-by-side | top-bottom | pip-top-left | pip-top-right | pip-bottom-left | pip-bottom-right | grid-2x2 | main-sidebar | three-up. mode 'fit' letterboxes each source in its cell, 'cover' fills the cell and crops the overflow (bias 0–1 picks the surviving part; 0.5/0.5 = centre). Replaces the clips' geometry animation. The clips should overlap in time on different tracks — a layout of clips that never share a frame arranges nothing visible.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline containing the clips.
  • clipIds (required) — array — Clips to arrange, in cell order.
  • layout (required) — string — Layout id (see description).
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • bias — object — Cover-crop bias 0–1 per axis (default 0.5/0.5).
  • mode — string — fit | cover (default fit).

  Hints: read-only no · destructive yes · idempotent yes · open-world no

• track_faces destructive

  Finds every FACE in a video clip and stores one independent path per person — position, size and head rotation (nod/turn/tilt) per frame. Queued, not immediate: poll list_face_tracks until state is 'ready'. Re-running on the same clip REPLACES its previous run and keeps the same trackId, so clips already bound stay bound. A person who leaves the shot and returns much later becomes a SECOND face on purpose — use merge_faces if they are the same person.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • clipId (required) — string — The video clip to scan, from get_timeline.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • name — string — Display name for the run.
  • holdSeconds — number — How long a vanished face is still 'the same face' when it comes back (default 1).
  • travelFaces — number — How far it may reappear from where it vanished, in face widths (default 0.8).
  • scoreThreshold — number — Detection confidence floor, 0.3–0.95 (default 0.6).

  Hints: read-only no · destructive yes · idempotent yes · open-world no

• list_face_tracks read-only

  Lists face-tracking runs and the people each one found: faceId, name, frame range, how many frames it was actually visible for, and how many gaps it has. Sample data is deliberately not returned. Use this to poll a queued run and to pick a faceId for bind_face.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • clipId — string — Only runs made from this clip.

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• merge_faces destructive

  Joins two or more faces from one run into a single person — for somebody who walked out of the shot and came back. REFUSED when any two of them appear on the same frame, since they cannot then be one person. The earliest face's id and name survive, so existing bindings keep working, and the time between sightings stays a gap for the clip's gap policy to handle.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • trackId (required) — string — The run, from list_face_tracks.
  • faceIds (required) — array — Two or more faceIds within that run.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive yes · idempotent yes · open-world no

• bind_face destructive

  Makes a clip FOLLOW a tracked face — the way to stick a mask, label or blur onto somebody's head. Movement is relative to where the face was first seen, so the clip keeps the framing it already had and only inherits the motion. Pick what to copy with channels, and what happens while the face is missing with gap. Pass trackId: null to unbind.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • clipId (required) — string — The clip that should follow the face.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • gap — string — While the face is missing: 'hold' stays put then jumps, 'hide' disappears, 'glide' travels smoothly to where it reappears. One of: hold | hide | glide.
  • faceId — string — Which person in that run.
  • smooth — number — Jitter filtering 0–1 (default 0.25).
  • trackId — string — The run; null unbinds this clip.
  • channels — object — What to copy: {position, scale, pitch, yaw, roll} booleans. Default position and scale on, the three rotations off.
  • maxGlideSeconds — number — Longest gap 'glide' will cross before hiding instead (default 1.8).

  Hints: read-only no · destructive yes · idempotent yes · open-world no

• organize_pool destructive

  THE MEDIA POOL'S SHAPE — bins (the pool's folders) and what sits in them. Use it to tidy a project: group the footage, the voice-over, the music and the graphics instead of leaving eighty files at the root. Bins are VIRTUAL and live in the project, not on disk. Moving an item between bins NEVER touches the file or its mediaRef, so it cannot break a clip's link or make the app ask the user to relocate anything — organise freely. actions: 'list' returns the bin tree and every item with the bin it shows in (call it first — you need the ids). 'create_bin' {name, parentBinId?}. 'rename_bin' {binId, name}. 'move_bin' {binId, toBinId} nests one bin inside another. 'delete_bin' {binId} — its children and items reparent, nothing is lost, no file is deleted. 'move_items' {itemKeys[], toBinId} moves files, timelines and motion compositions alike. Item keys: a media file's mediaRef, 'timeline:<id>', 'motion:<id>'. The Library root is the empty-string bin id. Bins that MIRROR a real folder (mirrorsFolder: true) can be RENAMED — a mirror is matched by the folder it tracks, not by its name — but they cannot be moved or deleted, because their place in the tree follows the folder. Move the items out instead, or make your own bins alongside.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • action (required) — string — list | create_bin | rename_bin | move_bin | delete_bin | move_items
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • name — string — Bin name (create_bin, rename_bin).
  • binId — string — Target bin (rename/move/delete).
  • toBinId — string — Destination bin for move_bin / move_items. Omit or '' = Library root.
  • itemKeys — array — move_items: mediaRefs / 'timeline:<id>' / 'motion:<id>' to move together.
  • parentBinId — string — Parent for create_bin. Omit or '' = Library root.

  Hints: read-only no · destructive yes · idempotent no · open-world no

• measure_loudness read-only

  MEASURE THE MIX — you cannot hear it; this is how you check it. Mixes the timeline's audio offline exactly as an export does (every audio track, nested timelines, fades, volume keyframes, clip audio effects, ducking, mute/solo) and measures it with EBU R128 / ITU-R BS.1770 — the same meter export loudness normalization uses. The mix is measured AS MIXED: an export's loudnessTargetLufs would then shift all of it by one gain. Returns {jobId}; poll get_job_status every few seconds. When done, `result` is {durationSec, integratedLufs, truePeakDbtp (4× oversampled, BS.1770-4), samplePeakDbfs, loudnessRangeLu, maxShortTermLufs, windows: [{fromSec, toSec, integratedLufs, maxShortTermLufs}] one per windowSec, sections?: [...same per requested section], speech: {source: transcript|none, speechSec, gapSec, mixDuringSpeechLufs, mixInGapsLufs, transcribedClips, untranscribedClips}, tracks: [{trackId, name, carriesSpeech, integratedLufs, samplePeakDbfs, duringSpeechLufs, inGapsLufs}]} (true peak is measured on the whole mix). SPEECH comes from transcript word timings mapped onto the timeline (trims, nests, speed): a GAP is a stretch of at least 0.6 s with no word within 0.25 s of it. `inGapsLufs` of a music/ambience track is the bed's level where nobody talks; compare it with the speech track's `duringSpeechLufs`. Without a transcript (speech.source 'none') the speech/gap fields are null — run start_transcription on the dialogue first. Integrated values are gated (−70 LUFS absolute, −10 LU relative), so silence does not drag them down; null means too short (under 0.4 s) or silent. Needs the project open in the editor tab you are paired with (PROJECT_NOT_OPEN otherwise).

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline to measure (what an export of it would sound like).
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • sections — array — Optional named spans to report on their own — e.g. the intro, each interview answer. Timeline seconds.
  • windowSec — number — Length of each entry in `windows` (3–600 s, default 10).

  Hints: read-only yes · destructive no · idempotent yes · open-world no

• duplicate_timeline

  Copy a timeline under a new name — a backup of the cut before a review round changes it, or a variant to try something on. A DEEP copy: new ids for the timeline, its tracks, clips and transitions; links between clips (A/V pairs, caption groups) are kept, inside the copy. Nested timelines are copied too by default (includeNested), so editing a nest in the original does not change the backup; motion compositions and media are referenced, not copied. The original is not touched and stays the active timeline; one undo removes the copy. Returns {timelineId, name, clips, nestedCopies?}.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline to copy.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • name — string — Name of the copy. Default '<name> copy'.
  • includeNested — boolean — Copy nested timelines too (default true). false = the copy shares them with the original.

  Hints: read-only no · destructive no · idempotent no · open-world no

• add_sample_media

  Frapea's SAMPLE MEDIA, for trying the editor without your own footage: a short talking clip with speech and pauses (for transcripts, captions, removing words or silences), a b-roll shot, a photo and a music bed. No API key, no download from elsewhere. Call with no `ids` to list them; call with `ids` (or `all: true`) to copy them into the project's media pool at the Library root — each result gives a mediaRef for add_clips. Adding one already in the project returns its existing mediaRef.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • all — boolean — Add every sample.
  • ids — array — Sample ids to add (talk, broll, photo, music). Omit to list them.

  Hints: read-only no · destructive no · idempotent yes · open-world no

• open_timeline

  Opens a timeline's TAB in the editor (the user can close tabs to keep the tab row short; a closed timeline is not deleted and every tool works on it as before). By default it does not change what the user is looking at; activate:true also shows it, like set_active_timeline. get_timeline reports `open` per timeline, as last saved.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline whose tab to open.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.
  • activate — boolean — Also show it (default false).

  Hints: read-only no · destructive no · idempotent yes · open-world no

• close_timeline

  Closes a timeline's TAB in the editor — hides it from the tab row, does NOT delete it: it stays in the media pool, and you (or anyone) can still edit, verify, nest and export it. When it was the one shown, the editor shows the next open tab, or none.

  Parameters:

  • projectId (required) — string — Project id from get_projects.
  • timelineId (required) — string — Timeline whose tab to close.
  • token (required) — string — The session token `connect` returned. Pass it on every call — it says which browser to drive.

  Hints: read-only no · destructive no · idempotent yes · open-world no

Publisher website

Measured by MCPYP probing. Untested fields are omitted, never estimated. Listing is always free; MCPYP never proxies calls or stores user tokens.