Feeling Search — Art Style Reference & Prompt Encyclopedia

cloud.xiximmPublished by xiximm.cloudOfficial publisherv0.2.6Updated Oct 30, 2026

One sentence in, the best-matching visual style out: art reference chart and prompt encyclopedia.

VerifiedStreamable HTTPWeb executableKnowledge & MemoryMedia & Design

Overview

AI-generated overview

Turns one colloquial sentence into matching visual art styles and the feeling words behind them, via a local lexicon and semantic engines or a remote endpoint.

What you can do
  • Analyze a sentence into feeling words and ranked visual styles, with the source layer reported
  • Look up a single feeling word: its styles, opposites, co-occurring words and whether it is a signature word
  • List the full feeling-word lexicon or the style set with definitions and sizes
  • Fetch one style in full: definition, generation prompt, words, era and region, plus reference image addresses
  • Check lexicon version, size and whether cloud quota is configured
When to use it
Useful when you want to go from a mood or scene description to concrete art-style references and prompt vocabulary, for illustration, design or image-generation briefs. The local package suits offline, private use; the remote endpoint avoids the download.
Requirements
Either the npm package feeling-search-mcp run with npx (first run downloads about 21 MB of lexicon and engines) or the remote streamable HTTP endpoint. No account or API key is needed. Optional settings include FS_MCP_URL or --endpoint, FS_MCP_TOKEN for a gated endpoint, --timeout, --remote and FS_DATA_ROOT.
Before you install
Tools are read-only and free by default, but the cloud fallback path is described as paid per call and only runs when explicitly allowed; without FS_QUOTA configured it should report that quota is needed rather than succeed. The remote endpoint sends your text off your machine, while the local package does not. Image entries marked as placeholders should not be treated as style samples.

Installation

In SourceWeft

  1. Open Feeling Search — Art Style Reference & Prompt Encyclopedia in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Use with an agent

Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.

Read https://sourceweft.com/mcp/cloud.xiximm%2Ffeeling-search/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.

Read the install guide the agent follows

Other MCP clients

Add this to your client's mcpServers config.

{
  "mcpServers": {
    "feeling-search-art-style-reference-promp": {
      "type": "http",
      "url": "https://xiximm.cloud/mcp"
    }
  }
}

README

feeling-search-mcp

MCP server for feeling-search — turn one colloquial sentence into the visual style that matches it best, plus the feeling words that led there. Every style carries its name, a one-line definition and the words that triggered it.

A commercial-grade art reference chart, art-style encyclopedia and art-style AI prompt dictionary, exposed over the Model Context Protocol.

"下雨天一个人走在便利店门口，冷白灯光"   → 感受词：星光 / 窗光 / 冷光 …   → 视觉风格：垂直式哥特 / 星图 / 托莱多画派 …

Install

Two shapes. Neither needs an account or an API key.

Local (default, recommended) — the package now ships the whole thing: 762-word lexicon, the 598-style index, and both semantic engines. It runs offline.

json
{  "mcpServers": {    "feeling-search": {      "command": "npx",      "args": ["-y", "feeling-search-mcp"]    }  }}

First npx run downloads ~21 MB (the two engines are 8 MB zh + 10 MB en). After that: lexicon hits answer in ~10 ms, whole-sentence queries in ~25 ms, all on your machine.

Remote (no download) — point the client at the endpoint:

json
{  "mcpServers": {    "feeling-search": {      "type": "streamableHttp",      "url": "https://xiximm.cloud/mcp"    }  }}

Or keep the local install but forward: npx feeling-search-mcp --remote. Remote answers in 0.76–1.64 s (network round trip + cold start), local in 10–90 ms, and local never sends your text anywhere.

Machine-readable copy of both variants: mcp.json.

Tools

ToolWhat it doesCost
word_analyzeOne sentence → the best-matching visual styles + the feeling words behind them. source tells you which layer answered (local lexicon / engine semantic model / cloud fallback)free, offline
word_lookupLook up a single feeling word: definition, near-synonyms, landing stylesfree
word_listBrowse / search the full feeling-word lexiconfree
style_getOne style in full: definition, parent style, member words, reference imagesfree
word_statusLexicon version, coverage — use it to check the data is freshfree

Every tool is read-only and costs nothing. The paid /v2/analyze cloud path is off by default on the public endpoint and only runs when a caller explicitly opts in.

Pricing status (2026-10-28): everything is free — no API key, no quota token. The upstream service runs free by default; paid 402 / x402 is only enabled if the operator explicitly sets FS_PAID=1. Only rate limits apply: 20 requests/minute per IP, and POST /v2/analyze is capped at 200 calls per visitor per day. Nothing in this bridge needs to change either way.

Want to know how much traffic the lexicon is getting? The upstream exposes a public GET /v2/stats telemetry endpoint (page views, unique visitors, referrers, AI crawler hits, no personal data) — see https://xiximm.cloud/docs/api.html.

Two layers, and which one answers

word_analyze only escalates when it has to:

LayerWhen it runsLatencyWhat it returns
Lexicon (source: "local")The sentence contains 2+ lexicon words~10 msExact feeling words, each with its landing styles
Semantic engine (source: "engine")The lexicon hits 0–1 words — i.e. it didn't really understand the sentence~25 msReal words from the 762-entry lexicon (cosine against 762 word vectors) + styles ranked against 598 style vectors
Cloud (source: "cloud")Only on the remote endpoint, and only with allow_cloud: true0.8–1.6 sPaid fallback

Two engines, picked by language:

  • Chinese — ask in whole Chinese sentences. Any Han character makes it the Chinese engine; don't mix, e.g. 有点 lonely 的雨夜 gets shredded by the Chinese tokenizer.
  • English — ask in whole English sentences. moody rainy night in old Hong Kong.
  • Force one with lang: "zh" | "en"; default is auto.

Negation is computed outside the model and is deterministic: 不要压抑 drops 压抑 and demotes every style that carries it, on both layers.

What's in the package

PathSizeWhy
data/src/words.json, data/standards/*.json~2 MBLexicon, 598 styles, aliases, version
data/standards/zh-engine/8.1 MBChinese ternary engine (wasm core + weights)
data/standards/ternlight/9.8 MBEnglish engine (wasm, weights embedded)
data/standards/idx/0.9 MB598 style vectors, per language, int8
lib/<100 KBTool definitions, engine loader, lexicon

Everything is generated from the upstream repo by deploy/pack-mcp.mjs, which also runs a --check drift comparison — the package can't silently fall behind the data.

Endpoint

PropertyValue
TransportStreamable HTTP (sse also accepted)
URLhttps://xiximm.cloud/mcp
Authnone — anonymous initialize and tools/list both succeed
Tools5 (word_analyze, word_lookup, word_list, style_get, word_status)
npm packagefeeling-search-mcp
MCP Registry namecloud.xiximm/feeling-search

Configuration

Env / flagDefaultMeaning
FS_MCP_URL / --endpointhttps://xiximm.cloud/mcpEndpoint to bridge to
FS_MCP_TOKENunsetSent as Authorization: Bearer <token> when the endpoint is gated
--timeout30000Per-request timeout in ms
--remoteoffForward to the endpoint instead of running locally
--data <dir> / FS_DATA_ROOTpackage data/Where the lexicon and engines live

stdout carries protocol messages only; all diagnostics go to stderr.

Reach & traffic

One command checks every channel this package is listed on — npm downloads, GitHub traffic and referrers, MCP Registry versions, directory availability, and open PRs:

bash
node scripts/traffic.mjs          # human-readable reportnode scripts/traffic.mjs --json   # machine-readable, for diffing over time

Requires the gh CLI to be logged in (GitHub traffic needs push permission on the repo). Mind the reporting lag: npm download stats appear 24–48h after the first publish, and GitHub traffic only covers the last 14 days.

referrers is the useful signal for directories — it shows which registry actually sends visitors. Directory sites themselves publish no view counts.

Related

License

MIT

Source: README.md at commit 7a6aedf

Tools

5

word_analyze

把一句描述解析成词与风格

把一句中文/英文口语描述解析成感受词与视觉风格。默认纯本地词典（免费、不联网）；本地一个词都没命中且允许时，才上云。返回 score 是相对值不是概率，styles[].matched 是可以直接引用给用户看的命中词。

Parameterslangtextrequiredquotawordsstylesallow_cloud

word_lookup

查一个词落在哪些风格

查一个感受词：落在哪些风格、对立词是什么、常一起出现的词、是否标识词。纯本地查表。

Parameterswordrequired

word_list

枚举感受词或视觉风格

列出全部感受词（带共享度）或视觉风格（带定义与词集规模）。调 analyze 之前先拉一次可避免名字拼错。

Parameterskind

style_get

取单个风格的定义与对应图片

按风格 id（或中文名）取一个风格：定义、生成用提示词、感受词、年代地域，以及它对应的图片（thumb 320 / full 960 绝对地址，带 kind/credit/license）。kind=work 是有出处的真作品，synthetic 是 AI 示意，placeholder 是占位图。

Parametersidrequired

word_status

查看词表版本与云端额度状态

返回词表版本、规模（词/风格/变种/挂靠率）以及是否配置了云端额度。

Takes no parameters.

Version history

2
  1. v0.2.6LatestOct 30, 2026
  2. v0.2.4Oct 29, 2026

Score

100% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 daysRequired25 pts
  • Auth declaredRequired10 pts
  • Tools listedRequired15 pts
  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts

Other Xiximm integrations

Create All Cg Save

Published by GitHub account xixi048

Locally run MCP server that finds offline single-player game saves and byte-patches CG unlock flags, with dry-run, backup and rollback.

STDIODesktop only
Files & StorageNot validated · the package type cannot run in the sandboxNov 2, 2026