MCP tool reference · #372 of 563 by publishers
semantic_search
Exposed by 11 servers from 10 publishers, read from the live servers by the Serverdex check; most of them sit in Media & Creative; 81.8% of them take a query parameter; the most-starred is DC Hub (3 stars).
semantic_search(query: string, limit?: integer, k?: number, q: string, top_k?: integer, …)Parameters
share = % of the 11 servers whose schema has it · required = % of those that mark it required| Parameter | Type | Share | Required | Usual description |
|---|---|---|---|---|
| query | string | 81.8% | 77.8% | Free-form question or phrase. |
| limit | integer | 45.5% | 0% | How many chunks to return. Default 6, max 15. Capped at 2 per document so one long file cannot take every slot. |
| k | number | 27.3% | 0% | Number of results, 1-15 (default 8) |
| q | string | 27.3% | 66.7% | Concept, question, or exact terms (≥2 chars) |
| top_k | integer | 27.3% | 0% | Maximum number of results to return (default: 5) |
| source_types | array | 18.2% | 0% | Restrict to these source types. "proposal" is a Dev Fund grant request; "github_item" is an ordinary issue or PR from any other indexed repo; "sdk_version" is a |
| auto_namespace | boolean | 9.1% | 0% | Auto-detect the namespace from the query text using text heuristics instead of using the `namespace` parameter. Returns results only from the matching domain (c |
| autonomyLevel | string | 9.1% | 0% | Filter by autonomy level |
| category | string | 9.1% | 0% | Filter by category slug |
| cohort | string | 9.1% | 0% | Filter by cohort |
| collection | — | 9.1% | 0% | — |
| corpus | string | 9.1% | 0% | Optional CSV of corpora: news_articles, deals, discovered_facilities, market_narratives (default: all) |
Source: the input schema each server returns for semantic_search, aggregated over the 11 index-tier servers; type and description are the values most servers use. Rebuilt every 6 hours.
Servers with semantic_search
11 of 11 · 10 publishers · 11 answering live · most-starred first, 3 per publisher| # | Server | Publisher | Category | ★ | Live | How it describes semantic_search |
|---|---|---|---|---|---|---|
| 1 | DC Hubazmartone67 · AI & Agents | azmartone67 | AI & Agents | 3 | yes | Use this when a data center or energy question is conceptual or fuzzy and keyword filters fall short — semantic (meaning-based) retrieval across DC Hub's industry news, M&A deals, the global facility map, per-market DCPI deep-dive narratives, press releases, announcements, permitting intel, construction permits, tax incentives and the capacity pipeline, ranked by relevance with citable source fiel |
| 2 | cachly — Explain it once. Every AI knows it.cachly-dev · Other | cachly-dev | Other | 2 | yes | Find cached entries that are semantically similar to a natural-language query. Read-only — no side effects. Returns an array of objects, each with: key, value, similarity_score (0–1), and namespace. Returns an empty array if no entries meet the similarity threshold. Requires OPENAI_API_KEY (or compatible provider) and CACHLY_VECTOR_URL; every tier includes a vector token. Embeddings are computed s |
| 3 | NEARpediaunitynodes · Documents & Knowledge | unitynodes | Documents & Knowledge | — | yes | Semantic (meaning-based) search over the NEAR knowledge base: docs.near.org and partner docs, gov.near.org threads, NEAR release notes, blog, the ecosystem catalog and GitHub issues of NEAR repositories. NEAR only: not for other blockchains, the general web or local files. Use this for conceptual how-does-X-work questions; use search for exact-term or name lookups, and get_doc to read a full page |
| 4 | web2mdio-oi-ai · AI & Agents | io-oi-ai | AI & Agents | — | yes | Semantic search over the user's saved conversions — matches by meaning, not exact words, and across languages. Use when they describe a page by topic rather than title or URL. Falls back to search_library for exact keyword matches. |
| 5 | Finekot APIdenyskot1986 · Media & Creative | denyskot1986 | Media & Creative | — | yes | Your own vector index behind the API key: upsert documents (chunked automatically), query in any language, get ranked chunks with metadata. Collections per key, 50 000 chunks. Upsert billed per 5 chunks. Cost €0.001 per query, charged to the API key in Authorization: Bearer. Parameters — op: query (default) | upsert | delete | list; collection: name (default 'default'); query: search text; top_k: |
| 6 | Cwi Discovery Enginecumulativewebinc · Media & Creative | cumulativewebinc | Media & Creative | — | yes | CALL THIS when the user describes music in plain words: "rainy night drive, nostalgic", "something that sounds like a heist". Returns catalog matches ranked by meaning rather than keywords. |
| 7 | RoboParts 机器人零部件兼容性lm203688 · IoT & Devices | lm203688 | IoT & Devices | — | yes | 用自然语言描述需求(如"人形机器人髋部高扭矩电机")做语义召回,返回最相近的零部件。语义索引由离线哈希 TF-IDF 向量(构建时预计算、零外发、零外部模型)在当前全量实体上生成,是兼容性判定之外的"发现"通道:当你不确定零件的确切型号或参数名时,用它比关键词子串匹配更稳。索引不可用时自动降级为关键词检索,并明确告知。 |
| 8 | Canton Ccpediaunitynodes · Crypto & Web3 | unitynodes | Crypto & Web3 | — | yes | PRIMARY tool for open-ended questions: how / why / what-is, troubleshooting a symptom ("why is my balance zero", "how do I fix X"), and locating config or setup steps. Conceptual/meaning-based search over the full Canton corpus (CIPs, docs, forum, mailing lists, proposals, blog, releases, ecosystem, foundation KB, YouTube) using vector+FTS hybrid retrieval with reranking. Canton-specific. Use this |
| 9 | Sniff MCPsniffscore · Health & Science | sniffscore | Health & Science | — | yes | Faceted hybrid + semantic-ranker search over the whole knowledge base (diseases, breeds, Scout discoveries). Use for fuzzy/thematic intent ('drug sensitivity in herding dogs', 'breeds prone to eye disease', 'genetically diverse breeds'). entity_type filters to 'disease'|'breed'|'discovery'. filters is an OData facet expression for cross-dimension queries, e.g. "breed_group eq 'herding' and cohort_ |
| 10 | Energy Podcast Digestm4rz910 · Media & Creative | m4rz910 | Media & Creative | — | yes | Passage search over episode summary sections AND verbatim transcripts: pass a concept or question ("who discussed interconnection queues") and get the closest passages, each with its episode, section, kind, rank, and score. `mode` picks the retriever: auto (the default — currently hybrid; `route` reports the signal), semantic (cosine pin), keyword (BM25 — exact acronyms, names, and numbers like PJ |
| 11 | Agentic.ai Directoryu00dxk2 · Files & Storage | u00dxk2 | Files & Storage | — | yes | Search for AI tools using natural language with AI-powered semantic matching. Best for conceptual queries like 'something that automates my email workflow'. Supports structured filters to narrow results (e.g., openSource + deploymentModel). For exact name/keyword searches, use search_listings instead. For comparing specific tools, use compare_listings. |
Stars from GitHub as stored in the Serverdex index (re-read every 7 days per server); live = answered the last daily check of its hosted endpoint.