Skip to main content
Glama

kelpo_expert_prepare

Espace Expert in one call, without conversation: prepare the whole scene from structured fields (who appears — catalogue actor, library person or a new person with 1-5 uploaded photos —, outfit, place, product photos, other images, place-visit presentation, what is said line by line in lines (speaker, exact words, tone) and what is seen in seen — or a free scenario —, voice, camera, background music, format). Upload the files first to this conversation (role unassigned, project_id), then call this once: the same operations and validations as the studio chat are applied atomically (status ready = recap ready; rejected = nothing saved, problem.section and problem.message say what to fix). Never generates nor charges: then quote and generate as usual, after the user's approval. Arguments: {"path": {"id": ""}, "body": {"version": 3, "dry_run": true, "kind": "facecam", "actors": [{"source": "catalog", "catalog_id": "fr-camille", "wardrobe": {"mode": "custom", "description": "pull beige simple"}}], "lines": [{"who": "Camille", "text": ""}], "location": {"mode": "named", "name": "cuisine lumineuse"}, "product_ids": [""], "camera": {"mode": "custom", "description": "selfie à bout de bras"}}}

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "path": "<path>",
      "body": "<body>"
    },
    "name": "kelpo_expert_prepare"
  }
}

Input Schema

NameRequiredDescriptionDefault
bodyYes
pathYes

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed3 schema fields changed
    • changedInput schema / $defs / ExpertActor / properties / photo_ids / description
      Previous value: -"upload: 1-5 identity photos (pending uploads of this conversation). library: the photos of that person used for this video, when they have more than the video takes (5 for one actor, 4 each from two, 3 for a place visit)."New value: +"upload: 1-5 identity photos (pending uploads of this conversation). library: only these photos of that person are used for this video (fewer is fine, e.g. to leave one out); empty = the first ones (5 for one actor, 4 each from two, 3 for a place visit)."
    • changedInput schema / $defs / ExpertInput / properties / product_ids / description
      Previous value: -"Product photos (pending uploads, 3 at most): shape, colours and branding are kept."New value: +"Product photos (pending uploads, up to 5): every angle the person has (front, sides, back…); actor, product and place photos make 9 images at most in all. Shape, colours and branding are kept."
    • changedInput schema / $defs / ExpertInput / properties / product_ids / maxItems
      Previous value: -3New value: +5
  2. Changed1 schema field changed
    • addedInput schema / $defs / ExpertLocation / properties / start_frame
      Added value: +{
      +  "default": false,
      +  "description": "photos with ONE place photo, or actor_background (that actor photo): the video opens exactly on it (its first frame), e.g. every shot must start on the person's image; false = only the place. Only for a single take, a photo at the video's format (16:9 ≥ 1344×768, 9:16 ≥ 768×1344) and Normal quality (HD only guides the opening).",
      +  "title": "Start Frame",
      +  "type": "boolean"
      +}
  3. Changed2 schema fields changed
    • changedInput schema / $defs / ExpertLocation / properties / catalog_id / description
      Previous value: -"catalog: the id of ONE view of kelpo_place_catalog (wide for two or three people, left/right seat for one person)."New value: +"Not offered on this account: leave empty."
    • changedInput schema / $defs / ExpertLocation / properties / mode / description
      Previous value: -"photos = uploaded place photos (asset_ids, the exact place); named = described in words (name); actor_background = the background of the first actor's photos; catalog = a Kelpo catalogue place (catalog_id, from kelpo_place_catalog); any = Kelpo chooses."New value: +"photos = uploaded place photos (asset_ids, the exact place); named = described in words (name); actor_background = the background of the first actor's photos; any = Kelpo chooses."
  4. Changed14 schema fields changed
    • addedInput schema / $defs / ExpertActor / properties / off_screen
      Added value: +{
      +  "default": false,
      +  "description": "Only their real voice is used, as the voice-over: they never appear and their photos are not used. Every actor off_screen = nobody on screen (a product, a place…).",
      +  "title": "Off Screen",
      +  "type": "boolean"
      +}
    • addedInput schema / $defs / ExpertInput / properties / aspect_ratio / anyOf
      Added value: +[
      +  {
      +    "enum": [
      +      "9:16",
      +      "1:1",
      +      "16:9"
      +    ],
      +    "type": "string"
      +  },
      +  {
      +    "type": "null"
      +  }
      +]
    • removedInput schema / $defs / ExpertInput / properties / aspect_ratio / default
      Removed value: -"9:16"
    • addedInput schema / $defs / ExpertInput / properties / aspect_ratio / description
      Added value: +"Format. Default: the format the conversation already has (16:9 once the 16:9 method of kelpo_get_youtube_flow is used), else 9:16. Always send \"16:9\" for YouTube long-form, a website or LinkedIn."
    • removedInput schema / $defs / ExpertInput / properties / aspect_ratio / enum
      Removed value: -[
      -  "9:16",
      -  "1:1",
      -  "16:9"
      -]
    • removedInput schema / $defs / ExpertInput / properties / aspect_ratio / type
      Removed value: -"string"
    • addedInput schema / $defs / ExpertInput / properties / recorded_audio_id
      Added value: +{
      +  "default": "",
      +  "description": "The person's own recording of the exact words (2-15 s, one clear voice, pending upload): the main person on screen says it, lip-synced; it replaces `lines` (the words are transcribed by Kelpo).",
      +  "pattern": "^(?:[a-f0-9]{32})?$",
      +  "title": "Recorded Audio Id",
      +  "type": "string"
      +}
    • changedInput schema / $defs / ExpertInput / properties / voice / description
      Previous value: -"actor = the voice of the speaking actor (catalogue, library or uploaded recording); generated = a generated voice, only when the user chose it."New value: +"actor = the voice of the speaking actor (catalogue, library or uploaded recording). generated is refused: Kelpo never makes a talking video without a real voice."
    • changedInput schema / $defs / ExpertInput / properties / voiceover_asset_id / description
      Previous value: -"Voice-over spoken with this recording (60 s minimum, pending upload) instead of an actor's or a generated voice."New value: +"Voice-over spoken with this recording (60 s minimum, pending upload) instead of an actor's voice."
    • changedInput schema / $defs / ExpertInput / properties / voiceover_voice / description
      Previous value: -"Voice of the off-screen narration (« Voix off : “…” » lines): the name of an actor of this video whose voice is used, or « generated »."New value: +"Voice of the off-screen narration (« Voix off : “…” » lines): the name of an actor of this video whose voice is used (never « generated »: refused)."
    • addedInput schema / $defs / ExpertLocation / properties / catalog_id
      Added value: +{
      +  "default": "",
      +  "description": "catalog: the id of ONE view of kelpo_place_catalog (wide for two or three people, left/right seat for one person).",
      +  "maxLength": 64,
      +  "title": "Catalog Id",
      +  "type": "string"
      +}
    • changedInput schema / $defs / ExpertLocation / properties / mode / description
      Previous value: -"photos = uploaded place photos (asset_ids, the exact place); named = described in words (name); actor_background = the background of the first actor's photos; any = Kelpo chooses."New value: +"photos = uploaded place photos (asset_ids, the exact place); named = described in words (name); actor_background = the background of the first actor's photos; catalog = a Kelpo catalogue place (catalog_id, from kelpo_place_catalog); any = Kelpo chooses."
    • changedInput schema / $defs / ExpertLocation / properties / mode / enum
      Previous value: -[
      -  "any",
      -  "named",
      -  "photos",
      -  "actor_background"
      -]New value: +[
      +  "any",
      +  "named",
      +  "photos",
      +  "actor_background",
      +  "catalog"
      +]
    • addedInput schema / properties / body / type
      Added value: +"object"
  5. First observed

TDQS

A4.4/5.0
Behavior5/5

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

Beyond annotations it discloses atomicity ('the same operations and validations as the studio chat are applied atomically'), the exact outcome semantics ('status ready = recap ready; rejected = nothing saved, `problem.section` and `problem.message` say what to fix'), and the cost boundary ('never generates nor charges'). No annotation contradiction: destructiveHint=false is consistent with the described all-or-nothing save.

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

Conciseness3/5

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

Purpose is front-loaded and every clause carries information, but the opening sentence is a roughly 100-word run-on dense with em-dash parentheticals that is hard to parse on first read. The trailing JSON example is long yet justified for a nested payload; the structure would benefit from splitting the field enumeration from the atomicity/status guidance.

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

Completeness4/5

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

For a complex nested write tool with no output schema, the description covers prerequisites, atomicity, rejection semantics, error-field locations and the cost/generation boundary. It stops short of describing the successful response payload beyond `status`/recap, but the essentials for a correct call are present.

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?

Top-level schema coverage is reported at 0% and the body is a deep nested object, so the description must compensate — and it does by naming the key fields and supplying a worked `Arguments` JSON showing `path.id`, `body.version`, `dry_run`, `kind`, actor with wardrobe, `lines`, `location` and `product_ids` wiring. The example is partial against the ~20 body fields, but the schema itself documents the remainder in detail.

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 states a specific verb and resource — 'prepare the whole scene from structured fields' — 'in one call, without conversation', and enumerates what the scene comprises (actors, lines/seen or scenario, location, product photos, voice, camera, music, format). This clearly separates it from a conversational studio flow and from the downstream generate/quote siblings it names. The purpose is unmistakable despite the dense prose.

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

Usage Guidelines4/5

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

It gives an explicit prerequisite sequence: 'Upload the files first to this conversation (role unassigned, project_id), then call this once', and an explicit boundary with the generation flow: 'Never generates nor charges: then quote and generate as usual, after the user's approval'. It lacks a direct statement of when to prefer this over a chat-driven or incremental flow, but the 'without conversation' framing implies it.

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

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources