Skip to main content
Glama

Treza

by io.github.treza-labs
OverviewDiscussionsAdmin

Assemble video

assemble_video
Destructive

Join clips the account already has into ONE finished video, in the order given, optionally over a voiceover (narrationUrl) and a music bed, which become separate tracks with the clips ducked under the voice, and with burned-in captions. The bed is a track from the library (musicUrl) or one composed for this cut from a description (musicPrompt), never both. One clip plus a track is how to put music or narration under a single video: "add music to my video" is clipUrls [that video] with a musicPrompt, and no pipeline to build. Every url must come from list_assets or a run's outputs in get_run. Stills hold 3 seconds each, or stretch so a longer voiceover plays in full; lengthSec in the result is how long the cut will run, and any warning in it should be passed on. Renders in the background: poll get_run with the returned pipelineId and runId. Free once the account has bought credits (the result carries charged: false), so never ask a paying account to top up before assembling; charged on a trial. Needs the pipelines:run and pipelines:write scopes.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
nameNoWhat to call the render, e.g. "Beach trip cut".
captionsNoTranscribe the finished edit and burn captions in. Only worth it when the clips or narration contain speech.
clipUrlsYesVideo or image urls in play order: two or more, or exactly one when narrationUrl, musicUrl or musicPrompt is set.
musicUrlNoMusic bed under the whole edit, beneath the clips' own sound and any narration: an audio url from list_assets. Leave it out when passing musicPrompt.
transitionNoHow each shot hands off to the next. Defaults to a hard cut.
musicPromptNoCompose the music bed for this cut instead of taking one from the library: genre, mood, tempo and instruments, e.g. "calm cinematic ambient, soft piano, gentle pads, no vocals". It sits and ducks exactly like a musicUrl bed and takes the same fades. Written by the Audio Generation node's default model (get_node_type music-gen), or its full-song model when the cut runs past 30 seconds; the result names the model as composedMusic. Leave it out when passing musicUrl.
orientationNo"vertical" is 1080x1920 for Shorts, Reels and TikTok, "horizontal" is 1920x1080, "match_first" (the default) keeps the first clip's shape. Shots that do not fill the frame are letterboxed, never cropped.
narrationUrlNoVoiceover laid over the whole edit. Everything else ducks under it.
audioFadeInSecNoFade the music bed (or the only track) up over this many seconds at the start. Applies to a composed bed too.
audioFadeOutSecNoFade the music bed (or the only track) down over this many seconds at the end; 2 to 4 reads as a deliberate ending. Narration is never faded when there is a bed.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed4 schema fields changed
    • changedInput schema / properties / audioFadeInSec / description
      Previous value: -"Fade the music bed (or the only track) up over this many seconds at the start."New value: +"Fade the music bed (or the only track) up over this many seconds at the start. Applies to a composed bed too."
    • changedInput schema / properties / clipUrls / description
      Previous value: -"Video or image urls in play order: two or more, or exactly one when narrationUrl or musicUrl is set."New value: +"Video or image urls in play order: two or more, or exactly one when narrationUrl, musicUrl or musicPrompt is set."
    • addedInput schema / properties / musicPrompt
      Added value: +{
+  "description": "Compose the music bed for this cut instead of taking one from the library: genre, mood, tempo and instruments, e.g. \"calm cinematic ambient, soft piano, gentle pads, no vocals\". It sits and ducks exactly like a musicUrl bed and takes the same fades. Written by the Audio Generation node's default model (get_node_type music-gen), or its full-song model when the cut runs past 30 seconds; the result names the model as composedMusic. Leave it out when passing musicUrl.",
+  "maxLength": 1000,
+  "type": "string"
+}
    • changedInput schema / properties / musicUrl / description
      Previous value: -"Music bed under the whole edit, beneath the clips' own sound and any narration."New value: +"Music bed under the whole edit, beneath the clips' own sound and any narration: an audio url from list_assets. Leave it out when passing musicPrompt."
  2. Added

TDQS

A4.8/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations only declare the safety profile (readOnly=false, destructive=true, idempotent=false), yet the description goes well beyond them: asynchronous background rendering with get_run polling via pipelineId/runId, billing semantics (charged:false for credit-holding accounts, 'never ask a paying account to top up', charged on trial), and required scopes (pipelines:run, pipelines:write). This is exactly the operational context an agent needs and cannot get from structured fields.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Dense but front-loaded, with the core action and ordering constraint first and the operational notes (polling, billing, scopes) last. It is long for a single paragraph and some clauses could be trimmed, but nearly every sentence carries decision-relevant information.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

There is no output schema, so the description compensates by naming the result fields that matter (pipelineId, runId, lengthSec, charged, composedMusic, warnings) and instructing the agent to relay warnings. For a 10-parameter asynchronous render tool, the coverage of inputs, lifecycle, and billing is complete enough to call it correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is already 100%, so the baseline is 3, but the description adds real semantics beyond the schema: the musicUrl/musicPrompt mutual exclusion, the single-clip-with-track case, stills holding 3 seconds or stretching to fit a longer voiceover, and the ducking behaviour of the bed. It stops short of covering every parameter (orientation, transitions are left to the schema), hence 4 rather than 5.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with a specific verb and resource ('Join clips ... into ONE finished video, in the order given') and immediately enumerates the variations (voiceover over, music bed under, captions burned in) so an agent can tell exactly what this tool produces versus siblings like run_pipeline or edit_asset.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It gives explicit when-to-use routing: single video with music/narration is done here with clipUrls plus musicPrompt and 'no pipeline to build,' whereas full pipelines belong to the sibling tools. It also names the alternatives for sourcing urls (list_assets, get_run outputs) and states the mutex rule (musicUrl vs musicPrompt, never both), leaving little to inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All Treza toolsInspect
Try in Browser
One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources