Civitai Mcp Ultimate logo

Civitai Mcp Ultimate

Organization
skillmds
civitai-mcp-ultimate

Ultimate MCP server for Civitai — search models, browse top images with prompts, download LoRAs/Checkpoints, analyze trends. Use when user asks about AI models, LoRAs, checkpoints, Stable Diffusion, Flux, image generation prompts, or Civitai.

Overview

Publisherskillmds
Repositoryskillmd
Skill namecivitai-mcp-ultimate
Stars
249
Forks
3
Bundled files
5
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 5 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by skillmds on GitHub. Read the source before you install it.

Installation

Install the Civitai Mcp Ultimate AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/skillmds/skillmd/tree/main/plugins/agents-mcp/skills/civitai-mcp-ultimate
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/skillmds/skillmd.git /tmp/skillmd
mkdir -p .claude/skills
cp -r /tmp/skillmd/plugins/agents-mcp/skills/civitai-mcp-ultimate .claude/skills/civitai-mcp-ultimate
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Civitai Mcp Ultimate in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Civitai Mcp Ultimate on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Civitai Mcp Ultimate is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Civitai MCP Ultimate — 14 Tools Reference

Quick Tool Map

GoalToolKey Params
Find models by namesearch_modelsquery (Meilisearch), types, base_model
Find models by creatorsearch_modelsusername (most reliable)
Batch fetch by IDssearch_modelsids=[123, 456]
Full model detailsget_modelmodel_id
Version details + hashesget_model_versionversion_id
Identify model from fileget_model_version_by_hashSHA256/AutoV2/CRC32/BLAKE3
Top checkpointsget_top_checkpointsbase_model, period
Top LoRAsget_top_lorasbase_model, nsfw
Browse images/videosbrowse_imagesmodel_id, content_type, tag, base_model
Trending images/videosget_top_imagessort, period, content_type, browsing_level
Example images for modelget_model_imagesmodel_id
Extract best promptsget_image_generation_datamodel_id
Download URLget_download_urlversion_id
Download commandsget_download_infomodel_id, comfyui_path

Common Recipes (Quick Reference)

Top image by likes yesterday, specific base model

get_top_images(sort="Most Reactions", period="Day", content_type="image", base_model="Flux.1 D", limit=1)

Top video by comments this week

get_top_images(sort="Most Comments", period="Week", content_type="video", limit=1)

Most popular SDXL model this month

search_models(base_model="SDXL 1.0", sort="Most Downloaded", period="Month", limit=1)

Top 5 images from the most popular Flux model today

search_models(types=["Checkpoint"], base_model="Flux.1 D", sort="Most Downloaded", period="Day", limit=1)
  -> get model_id
get_model_images(model_id=ID, limit=5)

Community images for a model (sorted by reactions)

browse_images(model_id=ID, sort="Most Reactions", period="Month", limit=5)

NOTE: model_id auto-resolves to the latest version. To see images from a specific version, use model_version_id directly:

get_model(model_id=ID)  # -> see all versions with IDs
browse_images(model_version_id=VERSION_ID, sort="Most Reactions", period="AllTime", limit=5)

Most popular NSFW image this month with prompt

get_top_images(sort="Most Reactions", period="Month", content_type="image", browsing_level="XXX", has_meta=true, limit=1)

Find a specific LoRA by name (e.g. hands fix)

search_models(query="hands fix", types=["LORA"], sort="Most Downloaded", limit=5)

NOTE: Text search uses Meilisearch (fast, accurate). Combining query + types + base_model works via Meilisearch! Falls back to REST API only when using REST-only params (ids, favorites, license filters).

Top 10 LoRA by downloads this month

get_top_loras(sort="Most Downloaded", period="Month", limit=10)

Anime images from ComfyUI, only with metadata

get_top_images(tag="anime", tools="ComfyUI", has_meta=true, period="Week", limit=5)

Only videos made on Civitai

get_top_images(content_type="video", made_on_site=true, period="Month", limit=5)

Only original works (no remixes)

get_top_images(originals_only=true, period="Week", limit=10)

XXX content only

get_top_images(browsing_level="XXX", sort="Most Reactions", period="Day", limit=5)
browse_images(browsing_level="X,XXX", content_type="video", period="Week", limit=3)

Image/Video Cache

All images and videos are automatically downloaded to ~/.civitai-mcp-cache/:

  • Images: 512px thumbnails (saves bandwidth)
  • Videos: original quality
  • Auto-cleanup: files older than 24h deleted on server startup
  • Local paths returned in output — use Read tool to preview images

Workflows

1. Find the Best LoRA for a Style

get_top_loras(base_model="Flux.1 D", period="Week", limit=10)
  -> pick model_id
get_model(model_id=ID)
  -> read trigger words, description, versions
get_image_generation_data(model_id=ID, limit=5)
  -> extract prompts, settings, LoRA weight
get_download_info(model_id=ID, comfyui_path="C:/ComfyUI/models")
  -> ready-to-paste curl/wget/PowerShell

2. Prompt Mining

get_top_images(sort="Most Reactions", period="Week", has_meta=true, limit=10)
  -> each image has: prompt, negative prompt, steps, CFG, sampler, seed, LoRAs

For a specific model:

get_image_generation_data(model_id=ID, sort="Most Reactions", limit=5)
  -> only images WITH metadata

3. Download Model to ComfyUI

get_download_info(model_id=858800, comfyui_path="D:/ComfyUI/models")

Returns: curl + wget + PowerShell commands, ComfyUI subfolder, all hashes, scan results.

4. Identify Unknown .safetensors File

get_model_version_by_hash(hash="E837144C55")  # AutoV2 hash

5. Compare Models by Popularity

get_top_checkpoints(base_model="Flux.1 D", period="Month", sort="Most Downloaded", limit=20)
get_top_checkpoints(base_model="SDXL 1.0", period="Month", sort="Most Downloaded", limit=20)

Enum Cheat Sheet

Model Types

Checkpoint, LORA, LoCon, TextualInversion, Hypernetwork, Controlnet, Poses, VAE, Upscaler, Wildcards, MotionModule, Workflows, Other

Base Models

CategoryValues
SD 1.xSD 1.4, SD 1.5, SD 1.5 LCM, SD 1.5 Hyper, SD 2.0, SD 2.1
SDXLSDXL 1.0, SDXL Lightning, SDXL Hyper
FluxFlux.1 S, Flux.1 D, Flux.1 Krea, Flux.1 Kontext, Flux.2 D, Flux.2 Klein 9B/4B
AnimePony, Pony V7, Illustrious, NoobAI, Anima
Z-ImageZImageBase, ZImageTurbo
OtherQwen, Chroma, HiDream, AuraFlow, Hunyuan 1, Kolors, PixArt a/E, Lumina
VideoCogVideoX, Hunyuan Video, LTXV, LTXV2, LTXV 2.3, Mochi, Wan Video 1.3B t2v, Wan Video 14B t2v, Wan Video 14B i2v 480p/720p, Wan Video 2.2 TI2V-5B/I2V-A14B/T2V-A14B, Wan Video 2.5 I2V/T2V

Sort

  • Models (Meilisearch): Most Downloaded, Highest Rated, Most Collected, Most Comments, Most Tipped, Newest, Oldest
  • Models (REST API): Highest Rated, Most Downloaded, Newest
  • Images/Videos: Most Reactions, Most Comments, Most Collected, Newest, Oldest

Period

AllTime, Year, Month, Week, Day

Browsing Level

PG (1), PG-13 (2), R (4), X (8), XXX (16) — comma-separated for multiple, e.g. "X,XXX"

Content Type

image, video

NSFW Levels (legacy)

None, Soft, Mature, X

Image Modifiers

has_meta, made_on_site, originals_only, remixes_only — boolean flags

Image Filters

tag (e.g. "anime"), base_model (e.g. "Flux.1 D"), tools (e.g. "ComfyUI"), techniques (e.g. "txt2img")

ComfyUI Folder Mapping

Checkpoint -> checkpoints/, LORA/LoCon -> loras/, TextualInversion -> embeddings/, Controlnet -> controlnet/, VAE -> vae/, Upscaler -> upscale_models/

Known Quirks

  1. search_models via REST API: query + types/base_model = empty results. Fixed in v0.2.0 — text search now uses Meilisearch which supports combining query with type/baseModel/tag/username filters. REST API fallback still has this limitation.
  2. search_models by username: Most reliable search method. Always prefer username="CreatorName".
  3. get_creators: Civitai endpoint is slow (30s+) and may return 500. Use search_models(username=...) instead.
  4. "Most Reactions" sort cursor bug: Returns null cursor, preventing pagination beyond page 1.
  5. License filters (allow_no_credit, etc.): Civitai API may return 400.
  6. modelId ignored on /images endpoint: Civitai API ignores modelId param — only modelVersionId works. Fixed in v0.2.1: browse_images, get_model_images, and get_image_generation_data now auto-resolve model_id to model_version_id (uses latest version). You can still pass model_version_id directly to target a specific version.
  7. browsingLevel: Bitmask system (1=PG, 2=PG-13, 4=R, 8=X, 16=XXX). Undocumented but stable.
  8. content_type/tag/tools/techniques: All undocumented API params but verified working 2026-03-24.
  9. get_tags: No NSFW filter available. Returns generic tags (character, anime, woman...) without model counts. Cannot query "top NSFW tags" — use browse_images(browsing_level="X,XXX") and inspect tags on returned images instead.

Environment Variables

CIVITAI_API_KEY     — Civitai API key (NSFW access + higher rate limits)
CIVITAI_MCP_LANG    — en (default) | ru
COMFYUI_MODELS_PATH — default ComfyUI models path for download_info
MEILISEARCH_KEY     — Meilisearch search-only key (optional, has built-in default)

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Civitai Mcp Ultimate AI skill do?

Ultimate MCP server for Civitai — search models, browse top images with prompts, download LoRAs/Checkpoints, analyze trends. Use when user asks about AI models, LoRAs, checkpoints, Stable Diffusion, Flux, image generation prompts, or Civitai.

Why use Civitai Mcp Ultimate on TypingMind?

Because you install it once and use it with any model. Civitai Mcp Ultimate is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Civitai Mcp Ultimate in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/skillmds/skillmd/tree/main/plugins/agents-mcp/skills/civitai-mcp-ultimate. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Civitai Mcp Ultimate?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Civitai Mcp Ultimate?

As many as you like. As long as a model supports skills, you can use Civitai Mcp Ultimate with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Civitai Mcp Ultimate AI skill free?

Yes. It is published on GitHub by skillmds under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇