
Listen to Sadhu MCP server
Which MCP servers are Listen to Sadhu's own, from the official registry and from asking its own domain, whether they answer, and what they offer. A program checked each of these, nobody copied them from a product page.
read 2026-10-31
- Registry nameapp.listentosadhu/corpusverified in the MCP registry
The registry name sits under listentosadhu.app. The registry only lets a publisher use that namespace after proving control of the domain through DNS or HTTP.
- Stateanswerschecked 2026-10-31
- Tools21
- Server calls itselflectorium-corpus-mcp
- Server version164e873d29
- Protocol version2025-06-18
On 2026-10-31 the server answered initialize and tools/list without a sign-in. First answer in 463 ms.
- Its tools do every use case we checked: listen to scripture lectures during a commute or walk; locate lectures by topic, author, or date.
- Checked on 5 of the 7 days it takes to judge reliability.
Once connected
- For "listen to scripture lectures during a commute or walk", the likely first call is track_list.
- For "locate lectures by topic, author, or date", the likely first call is track_list.
Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.
- INITE MCP scoreafter a weeka score needs 7 daily checks: 5 so far
- What an agent can do with it2 of 2 use cases
- Answeredmeasuring5 of the 7 daily checks it takes to print a share
- Median first answer469 ms
- Space in an agent's context≈ 3,358 tokensthe definitions of all its tools (21), which an agent carries on every turn
- Tool definitionsA · 98 of 100
- Typed results21 of 21 toolsdeclared outputSchema: results come typed, read once
- No description an agent can choose by: 2 of 21
What an agent can do with it
- can do · listen to scripture lectures during a commute or walk
- not sure · follow lecture transcripts while the audio plays
- can do · locate lectures by topic, author, or date
Each of the product's use cases, put to Jev against this server's tool list: can an agent do it with these tools?
Worth a look
None of the five warning signs in these definitions.
A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (25%), doing the product's job (25%), definitions (20%), context (10%), use (10%) and how readily a client connects (10%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →
author_listList all authors.
author_resolveFind a author by name (fuzzy).
document_getGet a document (commentary/prose_chapter/letter) by id.
document_listList documents at a reference / in a book. A verse's purport = document_list(source, tokens, kind:"commentary").
excerpt_prepareGenerate (or fetch, if cached) the audio clip for a lecture passage and return its playable URL. This is the excerpt player's Play action — it is called for you by the player UI on demand. Agents should use excerpt_render to show the…
excerpt_renderLet the user HEAR a lecture passage inline. Call this when the user wants to LISTEN to a specific moment in a track — e.g. from a search(types:["track"]) hit, pass its track_id plus start_ms/end_ms. Renders an inline audio player that…
location_listList all locations.
location_resolveFind a location by name (fuzzy).
media_renderShow/play an existing corpus VIDEO clip inline. Call this when the user wants to WATCH a clip found via search(types:["media"]) — pass the hit's media_id. Renders an inline video player.
searchSemantic + lexical search over the corpus. Returns verses, documents, track passages and titles matching a natural-language query, each with the id needed to fetch the full record (verse_id→verse_get, document_id→document_get…
source_getGet one book by id or code ("source_…" or "BG"/"БГ").
source_listList all books (Caitanya-caritāmṛta is three sources).
source_resolveFind a book by name (fuzzy): "gita", "бхагаватам", "CC Madhya".
track_getGet a track (lecture/conversation): metadata, cited references, transcript/pdf availability.
track_listList tracks, filterable by reference (source[+tokens] = tracks citing it), author/location/kind/date/lang. No filter ⇒ recent tracks (date desc).
transcript_windowRead a track's transcript around a time window (from a search track hit or track_list). Widen with pad_ms.
verse_getFetch a verse's raw fields for your own reasoning (NOT for display — to show a verse to the user, use verse_render). Returns original script + transliteration as line arrays; with lang, also the canonical translation inline plus an…
verse_listList the verses of a book / chapter. Skips .0 chapter summaries; collapses merged verses into one item with a covers span.
verse_renderShow a verse to the user. THE DEFAULT way to display/read a scripture verse: renders an inline card with the original script (centered), transliteration in the requested language, the word-by-word, and the translation — all at once. Use…
verse_synonymsFetch the word-by-word ("Synonyms") as a raw {word, meaning} array for one verse translation (data — verse_render already shows the word-by-word inside the card). Defaults to the canonical rendering; pass `kind` for an alternative.
verse_translationFetch the text of ONE translation of a verse (data, not a card — to display a verse use verse_render). Defaults to the canonical (kind=canonical); pass a `kind` from verse_get's `alternatives` manifest to fetch a specific rendering.
How to connect
Add the server to your MCP client config. Claude Desktop, Cursor and most other clients read this shape:
{
"mcpServers": {
"corpus": {
"type": "http",
"url": "https://mcp.listentosadhu.app/mcp"
}
}
}Add it
Claude Code, in a terminal
claude mcp add --transport http listentosadhu https://mcp.listentosadhu.app/mcpclaude.ai, ChatGPT and other clients: add https://mcp.listentosadhu.app/mcp as a custom connector in the settings.
Each line is something a program found when it went looking: the registry entry, the server's own answer, a spec that parses. A missing line means we did not find it, not that it does not exist.
Before we count this visit
Analytics and ad measurement tell us which pages and which of our ads do their job: Google Analytics and Google Ads, the Meta pixel, Yandex Metrica, and one first-party id of our own. Nothing is stored until you answer, and declining changes nothing about the site.