Skip to main content
Glama

About

LLMs are weak at the morphology and lexicon of Ethiopian and Eritrean languages. ethioMCP gives them tools backed by established linguistic repositories. Each repository is a connector, and adding a new one is a single folder (see docs/adding-a-connector.md).

Related MCP server: SolaGuard MCP Server

Connectors

hornmorpho: hltdi/HornMorpho

License: GPL-3.0. It runs as a separate process.

Tool

What it does

hornmorpho_analyze

Analyzes a word: lemma, root, POS, segmentation, UD features, UniMorph tags, transliteration, morphemes

hornmorpho_analyze_sentence

Analyzes a tokenized sentence and returns CoNLL-U

hornmorpho_generate

Generates word forms from a root (ስብር) or a lemma (ሰበረ), with features

hornmorpho_phonetic

Transliteration that shows gemination and epenthetic vowels

hornmorpho_languages

Supported languages: am, ti, om, tig

geezorg: geezorg/data

License: CC BY-SA 4.0.

Tool

What it does

geezorg_lookup

Dictionary search in either direction: Amharic (NMSU-Say, Science & Technology), Tigrinya and Ge'ez (Kidane Wolde Kifle)

geezorg_verb_paradigm

Armbruster Amharic verb tables: perfect, imperfect, jussive, gerund, and the other forms

geezorg_place_names

Ethiopian toponyms

geezorg_corpus_search

Keyword-in-context search of the POS-tagged WIC Amharic news corpus

geezorg_list_datasets, geezorg_browse, geezorg_read_file

Explore the rest of the archive

afrim: fodydev/afrim-data

The Ethiopic input methods of the afrim IME. License: none specified upstream.

Tool

What it does

afrim_transliterate

Romanised keyboard input to Ge'ez script (aityoPya → ኢትዮጵያ), with layouts am, gez, ethiopic, geezime; words found in the Amharic wordlist are flagged

afrim_keystrokes

Ge'ez script to the keystrokes that type it

afrim_lookup

Spelling check in a ~340k-form Amharic wordlist and an Ethiopian names list (no glosses)

afrim_geez_numeral

Integers to Ge'ez numerals (1966 → ፲፱፻፷፮)

afrim_ethiopian_date

Gregorian ↔ Ethiopian calendar, with Amharic month names

afrim_list_schemes

The layouts and their word lists

ethiomcp_connectors lists the connectors and whether each one is ready.

Install

git clone <this repo> && cd ethioMCP
uv sync
uv run ethiomcp setup                      # all connectors (HornMorpho: am,ti,om)
uv run ethiomcp setup hornmorpho -o langs=am
uv run ethiomcp setup afrim -o path=../afrim-data   # use a local clone instead of GitHub
uv run ethiomcp doctor

Downloaded data goes to a per-user cache directory, not to the project: ~/.cache/ethiomcp on Linux, ~/Library/Caches/ethiomcp on macOS, and %LOCALAPPDATA%\ethiomcp\Cache on Windows. Set ETHIOMCP_DATA_DIR to use another location.

Connectors that need their own Python environment get a venv inside that folder, with the Python version they declare. For example, HornMorpho runs on Python < 3.13, whatever version the server uses. uv downloads the interpreter if needed.

Use it

Claude Desktop (claude_desktop_config.json):

{
  "mcpServers": {
    "ethiomcp": {
      "command": "uv",
      "args": ["--directory", "/path/to/ethioMCP", "run", "ethiomcp", "serve"]
    }
  }
}

Claude Code:

claude mcp add ethiomcp -- uv --directory /path/to/ethioMCP run ethiomcp serve

ChatGPT (developer mode or custom connectors) needs an HTTP endpoint:

uv run ethiomcp serve --transport streamable-http --host 0.0.0.0 --port 8000
# expose https://<host>/mcp, e.g. with a reverse proxy or tunnel

To enable only some connectors: ETHIOMCP_CONNECTORS=geezorg uv run ethiomcp serve.

The first HornMorpho call in a session loads the language model (~20–60 s). Later calls are fast, because the worker stays alive.

Development

uv run ruff check && uv run ruff format --check
uv run pytest                    # unit tests, no network
uv run pytest -m integration     # real data (run `ethiomcp setup` first)
npx @modelcontextprotocol/inspector uv run ethiomcp serve

License

The ethioMCP code is MIT. Each connector's data and libraries keep their upstream license, which is reported with every tool result. HornMorpho (GPL-3.0) is never imported into the server process.

Available Tools

19 tools
afrim_ethiopian_dateA

Convert a date between the Gregorian and the Ethiopian calendar. date: DD/MM/YYYY, DD.MM.YYYY, DD-MM-YYYY (2-digit years are 20YY) or YYYY-MM-DD. calendar_of_input: 'gregorian' (default) or 'ethiopian'. Returns both dates, the Ethiopian month name, the year in Ge'ez numerals, and formatted strings such as '6 ሕዳር 2019' and 'ሕዳር 6 ቀን 2019 ዓ.ም.'.

Inputs: date (required), calendar_of_input. Returns: object

afrim_geez_numeralA

Write a positive integer (up to 13 digits) with Ge'ez numerals, e.g. 1966 -> ፲፱፻፷፮, 10000 -> ፼, 2018 -> ፳፻፲፰.

Inputs: number (required). Returns: object

afrim_keystrokesA

Give the keystrokes that type Ge'ez-script text with an afrim keyboard layout (the reverse of afrim_transliterate), e.g. ኢትዮጵያ -> 'aityoPya'. Each character uses the shortest key sequence. Returns the keystroke string and a per-character breakdown. scheme: am | gez | ethiopic | geezime.

Inputs: text (required), scheme. Returns: object

afrim_list_schemesA

List the input-method schemes (keyboard layouts) with their language, number of keys and word lists. Scheme ids: am, gez, ethiopic, geezime.

Inputs: no named parameters. Returns: object

afrim_lookupA

Search the Amharic wordlist (~340k inflected word forms crawled from the web) and the Ethiopian personal-names list. Use it to check that an Amharic word exists and how it is spelled; there are no glosses. A Ge'ez-script query (e.g. 'ኢትዮጵያ') searches the words; a romanised query (e.g. 'aityoPya') searches the keyboard spellings. Ranking: exact, case-insensitive (romanised only), prefix, then substring. dataset: 'dictionary' or 'names' to restrict the search.

Inputs: query (required), limit, dataset. Returns: object

afrim_transliterateA

Convert romanised keyboard input to Ge'ez script, as the afrim input method types it (greedy longest key match). The 'am' layout is SERA-like and case-sensitive: consonant + e u i a ie o (ሰ=se ሱ=su ሲ=si ሳ=sa ሴ=sie ሶ=so), a bare consonant is the 6th form (ስ=s), capitals are other consonants (ጠ=Te ጨ=Ce ጰ=Pe ሸ=xe ቀ=qe), and a vowel standing alone takes an 'a' prefix (አ=ae ኢ=ai እ=a). Examples: 'selam' -> ሰላም, 'aityoPya' -> ኢትዮጵያ, 'biet' -> ቤት. Punctuation: ':' -> ፡, '::' or '.' -> ።, ',' -> ፣. Use afrim_keystrokes to find the keys for a Ge'ez word.

scheme: am (default) | gez | ethiopic | geezime. use_wordlist: with 'am', a word that is in the Amharic wordlist or the names list is replaced by its listed spelling (useful where the keyboard map is ambiguous). Each word reports whether it was found there. Returns the text and a per-word breakdown.

Inputs: text (required), scheme, use_wordlist. Returns: object

ethiomcp_connectorsA

List the connectors on this server: what each one is, where its data comes from, its license, and whether it is ready to use.

Inputs: no named parameters.

geezorg_browseB

List the files and folders at a path of the archive, for example 'amharic/dictionaries'. The archive also holds abbreviations, plurals, homonyms, gemination lists, parallel Amharic-English news and Sabaean/Falasi glossaries that have no dedicated tool.

Inputs: path. Returns: object

geezorg_list_datasetsA

List the structured datasets (dictionaries, verb paradigms, place names) that the geezorg tools search, with their language and source files.

Inputs: no named parameters. Returns: object

geezorg_lookupA

Look up a word in the dictionaries.

A query in Ge'ez script (e.g. 'ቤት') searches headwords. A Latin-script query (e.g. 'house') searches the English glosses. Results are ranked exact match, then prefix, then whole word, then substring. lang: am | ti | gez filters by language. dataset: a dataset id from geezorg_list_datasets.

Inputs: query (required), lang, limit, dataset. Returns: object

geezorg_place_namesC

Search Ethiopian place names, in Amharic script or in romanised form.

Inputs: query (required), limit. Returns: object

geezorg_read_fileA

Read lines [offset, offset+limit) of a text file in the archive (.txt/.csv/.json/.tex/.xml). Use geezorg_browse to find paths.

Inputs: path (required), limit, offset. Returns: object

geezorg_verb_paradigmB

Return the conjugation table of an Amharic verb from the Armbruster list: simple and compound perfect, contingent, imperfect, jussive, imperative, gerund, infinitive, participle, translation, verb type. Give the 3rd person masculine perfect in Ge'ez script (e.g. 'ሰበረ'), or an English translation (e.g. 'break'). Gemination is marked with U+135F.

Inputs: verb (required), limit. Returns: object

hornmorpho_analyzeA

Analyze one word in Ge'ez script (e.g. 'የቤታችን').

Returns every analysis, with its lemma, root, part of speech, segmentation ('seg', stem between < >), Universal Dependencies features ('udfeats'), UniMorph tags ('um'), a transliteration ('translit', gemination shown) and per-morpheme details. guess=True guesses analyses for unknown words. include_feats=True also returns HornMorpho's full feature structure, which shows the feature names that hornmorpho_generate accepts. lang: am | ti | om | tig.

Inputs: word (required), lang, guess, include_feats. Returns: object

hornmorpho_analyze_sentenceA

Analyze a tokenized sentence (words separated by spaces, punctuation split off). Returns a CoNLL-U annotation with lemmas, UPOS, features and morpheme segmentation (clitics become separate tokens). Full dependency trees are not predicted. lang: am | ti.

Inputs: sentence (required), lang. Returns: object

hornmorpho_generateA

Generate word forms from a root or lemma in Ge'ez script.

root: a consonantal root such as 'ስብር', or a lemma such as 'ሰበረ' (its root is found automatically). features: a HornMorpho feature structure, for example '[sb=[+p1,+plr]]' for a 1st person plural subject. Run hornmorpho_analyze with include_feats=True on a similar word to see the feature names. pos: 'v' (verb), 'n' (noun), 'vn' (verbal noun), 'cop'. Without pos and features, the citation forms of each derived POS are returned.

Inputs: root (required), pos, lang, features. Returns: object

hornmorpho_languagesA

List the languages HornMorpho supports, with their codes for the other hornmorpho tools.

Inputs: no named parameters. Returns: object

hornmorpho_phoneticA

Phonetic transliteration of a Ge'ez script word, showing gemination and the epenthetic vowels the script does not write. One entry per distinct reading (e.g. ሰበረ -> səbbərə or səbərə).

Inputs: word (required), lang. Returns: object

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.6/5.0

    Scored across 19 tools

    Disambiguation4/5

    Most tools have clearly distinct purposes across the afrim, geezorg, and hornmorpho connectors. Some potential confusion exists between afrim_lookup and geezorg_lookup, and between geezorg_verb_paradigm and hornmorpho_generate, but the descriptions clarify dataset and operation differences.

    Naming Consistency4/5

    Tool names consistently use snake_case with connector prefixes (afrim_, geezorg_, hornmorpho_, ethiomcp_), which makes grouping predictable. Most names follow a verb/noun operation pattern, though a few list-style tools are noun phrases rather than verb-led.

    Tool Count4/5

    The 19 tools are spread across four distinct connectors, so the count is broad but reasonable for a multi-domain Ethiopian language toolkit. Each tool appears to serve a specific operation rather than duplicating another.

    Completeness4/5

    The surface covers transliteration, keyboard mapping, dictionary lookup, morphology analysis/generation, corpus search, dates, numerals, and archive browsing/reading. Minor gaps exist, such as no write/update operations for archive files and no translation or audio features, but core workflows are well covered.

    Maintenance

    ActivityMaintained
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      D
      maintenance
      Provides access to the Database of Icelandic Morphology (BÍN) to look up word inflections, variants, and lemmas. It enables LLMs to answer grammatical questions and retrieve specific Icelandic word forms using standard grammatical tags.
      3
      4
      MIT
    • A
      license
      A
      quality
      D
      maintenance
      Provides comprehensive Biblical research tools including scripture lookup, interlinear Greek/Hebrew data, and Strong's concordance within a Protestant theological framework. It enables AI applications to perform full-text biblical searches, topical studies, and cross-referencing using authoritative theological data.
      4
      MIT
    • A
      license
      A
      quality
      B
      maintenance
      Provides AI agents with local Dutch spelling and word validation using OpenTaal/hunspell, plus detailed word information from woordenlijst.org. It enables checking full texts or individual words and retrieving linguistic details for Dutch writing tasks.
      5
      6 npm
      MIT