dev.tanod/ml

1 of 18 tools are described as changing something (creating, editing, deleting, sending or running). For 4 the reading was uncertain, so no mark is shown.

• embed_textson 1 other server

  “mlpeek: Create text embeddings.”

  Publisher declares: read-only yes · destructive no · idempotent yes · open-world no

  5 inputs, 1 required

  • textsarrayrequired“1-64 texts, each at most 8,000 characters; each is truncated at 512 tokens (reported per text).”
  • modelstring“small-en = BAAI/bge-small-en-v1.5 (English); multilingual = intfloat/multilingual-e5-small (~100 languages).”
  • normalizeboolean“L2-normalise (cosine similarity = dot product).”
  • encodingstring“base64 = little-endian float32 bytes.”
  • input_typestring“Retrieval prefix: query / passage (e5: 'query: ' / 'passage: '; bge: its query instruction / nothing).”
• embed_multimodalon 1 other server

  “mlpeek: Create multimodal embeddings: images and text in one vector space.”

  Publisher declares: read-only yes · destructive no · idempotent yes · open-world no

  5 inputs, 1 required

  • itemsarrayrequired“1-16 items, each {text}, {image_url} or {image_base64}; every vector lands in one shared space.”
  • dimensionsinteger“768 (default), 512, 256 or 128: the leading values, re-normalised (Matryoshka).”
  • input_typestring“Task prefix for text items (images get none): query (search query), document (search corpus), question, fact_check, code_query, classification, clustering, sim…”
  • image_detailstring“Vision tokens per image: low (70), standard (140) or high (280, slowest).”
  • encodingstring“base64 = little-endian float32 bytes.”
• rerank_documentson 4 other servers

  “mlpeek: Rerank up to 100 documents against a query.”

  Publisher declares: read-only yes · destructive no · idempotent yes · open-world no

  3 inputs, 2 required

  • querystringrequired“The query (at most 2,000 characters).”
  • documentsarrayrequired“1-100 documents, each at most 4,000 characters.”
  • top_kinteger or null“Keep the best k (default: all, ranked).”
• text_similarityon 2 other servers

  “mlpeek: Score the semantic similarity of one or up to 50 text pairs.”

  Publisher declares: read-only yes · destructive no · idempotent yes · open-world no

  4 inputs

  • astring or null“First text of a single pair (at most 4,000 characters).”
  • bstring or null“Second text of a single pair.”
  • pairsarray or null“Alternative to a / b: 1-50 {a, b} pairs.”
  • modelstring“small-en (English) or multilingual.”
• extract_entitieson 6 other servers

  “mlpeek: Extract named entities: people, organisations, places, dates, money, ...”

  Publisher declares: read-only yes · destructive no · idempotent yes · open-world no

  2 inputs, 1 required

  • textstringrequired“English text, at most 20,000 characters.”
  • labelsarray or null“Only these entity types (default: all 18 OntoNotes types).”
• classify_zero_shoton 1 other server

  “mlpeek: Classify a text into 1-10 of your labels.”

  Publisher declares: read-only yes · destructive no · idempotent yes · open-world no

  4 inputs, 2 required

  • textstringrequired“English text, at most 2,000 characters.”
  • labelsarrayrequired“1-10 unique candidate labels, each at most 100 characters.”
  • multi_labelboolean“Score each label on its own instead of making them compete.”
  • hypothesis_templatestring“Must contain {} exactly once.”
• text_sentimenton 4 other servers

  “utilpeek: Analyze the sentiment of an English text.”

  Publisher declares: read-only yes · destructive no · idempotent yes · open-world no

  2 inputs, 1 required

  • textstringrequired“The text (at most 200,000 characters).”
  • per_sentenceboolean“Also score each sentence (default false: whole text only).”
• text_keywordson 5 other servers

  “utilpeek: Extract the keywords and keyphrases of a text.”

  Publisher declares: read-only yes · destructive no · idempotent yes · open-world no

  4 inputs, 1 required

  • textstringrequired“The text (at most 200,000 characters).”
  • languagestring“Stop-word language: de, en (default), es, fr, it, nl or pt.”
  • top_ninteger“Keyphrases to return, best first (1-100, default 10).”
  • max_wordsinteger“Longest keyphrase, in words.”
• text_summarizeon 10 other servers

  “utilpeek: Summarize a text by picking its key sentences.”

  Publisher declares: read-only yes · destructive no · idempotent yes · open-world no

  4 inputs, 1 required

  • textstringrequired“The text (at most 200,000 characters).”
  • languagestring“Stop-word language: de, en (default), es, fr, it, nl or pt.”
  • sentencesinteger“Sentences to pick (ignored when ratio is set).”
  • rationumber or null“Pick this share of the sentences instead (at most 100).”
• detect_languageon 19 other servers

  “utilpeek: Detect the language of a text.”

  Publisher declares: read-only yes · destructive no · idempotent yes · open-world no

  1 input, 1 required

  • textstringrequired“The text (at most 20,000 characters).”
• transcribe_audioon 20 other servers

  “mlpeek: Speech to text and subtitles: transcribe audio to text, timed segments, SRT and WebVTT.”

  Publisher declares: read-only yes · destructive no · idempotent yes · open-world yes

  7 inputs

  • urlstring or null“Public http(s) URL of the audio file (at most 25 MB; private, internal and IP-literal hosts are refused); or use file_base64.”
  • file_base64string or null“The audio file as base64 (at most 25 MB decoded); or use url.”
  • languagestring or null“Optional spoken language as an ISO 639-1 code (en, es, fr, de, tl, ...).”
  • translate_to_englishboolean“true: the text, segments, srt and vtt come out in English from any spoken language (Whisper translate task); `language` is still the spoken language.”
  • speakersboolean“true: speaker labels (diarization).”
  • num_speakersinteger or null“With speakers: the number of speakers when you know it (1-8); otherwise it is estimated.”
  • chaptersboolean“true: a `chapters` list with the start and end time, a short title and a one-line gist of each part (about one per three minutes, at most six), written by a sm…”
• transcribe_podcast_episodeon 1 other server

  “mlpeek: Transcribe a podcast episode from its RSS feed: a window of up to 10 minutes as text, timed segments, SRT.”

  Publisher declares: read-only yes · destructive no · idempotent yes · open-world yes

  9 inputs, 1 required

  • feed_urlstringrequired“Public http(s) URL of the show's RSS feed (find it with /v1/podcast/search).”
  • guidstring or null“Pick the episode by its feed guid.”
  • episode_numberinteger or null“Pick the episode by its itunes:episode number.”
  • seasoninteger or null“With episode_number: the itunes:season number.”
  • episode_titlestring or null“Pick the newest episode whose title contains this text (case-insensitive; an exact title wins).”
  • start_snumber“Where the window starts, in seconds from the beginning of the episode (default 0).”
  • duration_sinteger“Window length in seconds, 5-600 (default 600, the 10-minute limit of one call).”
  • languagestring or null“Optional spoken language as an ISO 639-1 code (default: detected).”
  • translate_to_englishboolean“true: text, segments, srt and vtt in English (Whisper translate).”

Each quoted sentence is the first sentence of the publisher’s own description. The kind of tool and the “changes data” mark are our AI model’s reading of that description, not something we tested: two differently worded runs had to agree, otherwise no label is shown, and nobody has checked them by hand. A tool without the “changes data” mark is described as only reading, searching or calculating, or the two runs did not agree. No tool was called. Read the publisher’s full description before relying on it. “Publisher declares” lists the hints the server itself sends with the tool (MCP annotations: read-only, destructive, idempotent, open-world); they are the publisher's claims, not checked by us. Inputs are the parameters in the tool's own input schema: name, declared type, whether it is required, and the first sentence of its description. Read 2026-10-23.

The registry verifies a package once, when it is published, and does not re-check it. A difference below is about the listing, not a judgement of the publisher.