FrameThrower

search_animation_frames

Read-only

Find frames in FrameThrower's ANIMATION library — animated series and films only, never live-action (mostly anime, plus Western animation; the whole of every title, not a sample) for a scene, a moment or a CHARACTER LOOK, described in plain words: "a girl on a rooftop at dusk watching the city lights", "two kids running through a summer field", "a tall man with a scarred face and a long black coat, rain". Narrow by title, genre, theme, style (80s cel, watercolour, neon night…), studio, movie/series and year. Each frame has a caption of what is in it, a 640px still for looking at, the full-size still, the frame id and a link. At most 3 frames per title unless per_title says otherwise.

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": {
      "query": "<query>"
    },
    "name": "search_animation_frames"
  }
}

Input Schema

Name	Required	Description
`kind`	No
`nsfw`	No	false = leave out frames flagged as adult (included by default)
`clips`	No	true = also return each frame's clip and sequence videos (the call costs 8 credits). Ask only when the user wants to see it move.
`genre`	No	AniList genre: Action, Adventure, Comedy, Drama, Fantasy, Horror, Mecha, Mystery, Psychological, Romance, Sci-Fi, Slice of Life, Sports, Supernatural, Thriller… (several, comma-separated = all)
`limit`	No	default 12
`query`	Yes	The scene or character in plain words, as a description of one frame
`style`	No	The title's look, read off its frames (several, comma-separated = all): hand-drawn cel, digital 2D, 3D CGI, cel-shaded 3D, stop motion, mixed 2D/3D, 70s look, 80s look, 90s look, 2000s look, modern look, pastel, muted, vivid, dark, warm, cool, monochrome, clean lines, thick outlines, sketchy lines, painterly, watercolour, flat minimal, detailed backgrounds, film grain, realistic proportions, moe big-eyed, chibi, western cartoon, gritty mature, bishonen, soft natural light, dramatic light rays, glow and bloom, neon, high-contrast shadows
`theme`	No	AniList theme tag: Cyberpunk, Post-Apocalyptic, Iyashikei, Coming of Age, Time Travel, Magic, Space, Urban Fantasy…
`title`	No	Only inside this series / film (slug or name)
`offset`	No
`studio`	No	Animation studio, e.g. "Studio Ghibli", "Madhouse", "Kyoto Animation", "MAPPA"
`year_max`	No
`year_min`	No
`per_title`	No	Max frames from one title (default 3)
`exclude_titles`	No	Slugs to leave out (e.g. titles already used)

Schema Changelog

Changes observed during successful MCP inspections.

AddedNov 3, 2026

TDQS

A4.3/5.0

Behavior4/5

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

Annotations already declare readOnlyHint and openWorldHint, so safety is covered; the description contributes the return payload (caption, 640px still, full-size still, frame id, link) and the per-title diversity cap of 3 — both non-obvious behaviors an agent would otherwise not expect. It omits credit economics and pagination ceilings, though those live in the schema.

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?

Front-loads the resource and its exclusivity before any parameters, and every clause carries content — the corpus scope, the query style, the filters, the return shape. It is dense and parenthetical-heavy in places, making it a single long breath rather than cleanly separated sentences, but nothing is filler.

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 15-parameter search tool with no output schema, the description carries the important loads: what is searchable, how to phrase a query, what comes back per frame, and the default result diversification. Only edge concerns like pagination bounds and credit cost are left to the schema instead of being surfaced here.

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?

Maps most filter dimensions to their purpose in prose (narrow by title, genre, theme, style, studio, movie/series and year) and adds interpretive examples for style such as 80s cel and watercolour, plus the per_title override. With 73% schema coverage the baseline is 3, and this description earns a step above it by explaining intent rather than restating field names.

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?

States a specific verb plus resource — find frames in FrameThrower's ANIMATION library — and immediately scopes it (animated series and films only, never live-action). That single clause separates it from search_commercial_shots and frames the relationship to find_animation_titles without opening a schema.

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?

Gives a clear use case with three concrete plain-language query examples and states the corpus boundary (whole of every title, not a sample), which tells the agent when this is the right retrieval tool. It never names its closest siblings, search_frames or similar_animation_frames, so the agent must infer which one handles a duplicate/visual-similarity request.

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

All FrameThrower tools Inspect

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.