_ registry / mcp streamable-http · checked 22m ago

soapbox-faith

https://foyekanoxpnkydoibaas.supabase.co

Registry code: 2df2c613bddd3f4c

api record

Grounded, public-domain (KJV) Scripture tools. Call verify_scripture before presenting any Bible quote to avoid hallucinated/misattributed verses.

endpoint
https://foyekanoxpnkydoibaas.supabase.co/functions/v1/faith-mcp
protocol
streamable-http ·2025-06-18
authentication
none observed
public key
none — nobody has proven they own this listing · is it yours? claim it
karma
0 · newcomer
_ is it live, free and safe measured by this hub
Is soapbox-faith live?
Yes — it answered the hub's last check (checked 22m ago). It answered 100% of checks over the last 30 days.
Is soapbox-faith free to use?
Partly — some of its tools are open, others need a key or payment.
What tools does soapbox-faith have?
25 tools: search_sermons, submit_prayer_request, sermon_study_guide, lookup_strongs, compose_prayer_request, give_to_church, ask_ora, pay_with_x402, ….
Is soapbox-faith safe to connect?
The hub found no text in its card or tool descriptions aimed at the agent reading them. It measures what the server answers, not its code — grant it only the access its tools need.
reachable
live
uptime, 30 days
100%

90 days 99.8%· all time 99.8%

latency
204ms

last good check

priced tools
0

of 25 tools

_ answered our checks, 90 days 466 checks · signed record
  • degraded → live
  • degraded → live
  • live → degraded· timeout after 20000ms
  • unknown → live
_ what it is for
topic
knowledge & reference
used for
  • verify bible quote
  • get daily devotional
  • find churches
  • search sermons
takes → gives
text → text, data, audio
tools
21 reads2 changes data2 moves money
_ usage and payments 30 days

Calls placed through this hub's router, from its own receipts. Every caller and every payer counts the same; the chain total is counted from three payers.

accounts
0

through this hub

calls served
0

successful

paid through this hub
0 USDC

what callers paid

_ what it can do 25 tools
3 open3 auth-required 19 never probed 6 of 25 classified

Price is per tool, not per server. An agent whose handshake is open can hold tools that demand a key or a payment, and one figure for the whole agent sends callers into a wall.

  • browse_catalog reads open 5h ago

    Browse purchasable faith-content products on SoapBox (currently consented sermons; reading-plan and devotional bundles coming). Returns products with id, title, church, and price. Then pay for one with pay_with_x402 (USDC on Base). Optional church_id and text filter. No key required.

    mcp-tool

    {
  "type": "object",
  "properties": {
    "q": {
      "type": "string",
      "examples": [
        "romans",
        "advent"
      ],
      "description": "Optional case-insensitive text filter matched against product titles."
    },
    "limit": {
      "type": "integer",
      "default": 25,
      "maximum": 100,
      "minimum": 1,
      "description": "Maximum number of products to return."
    },
    "church_id": {
      "type": "string",
      "format": "uuid",
      "examples": [
        "b3095717-5cc0-4213-8964-a01a5ee6b250"
      ],
      "description": "Optional SoapBox community id to scope the catalog to a single church."
    }
  }
}
    arguments 28 lines
  • get_daily_devotional reads open 5h ago

    Build a season-aware daily devotional for a date: the liturgical season and colour, a KJV passage chosen for that season, three reflection QUESTIONS, and a prompt for the member to pray in their own words. No key required. IMPORTANT, two things this is not. (1) It is NOT the lectionary: the reading is SoapBox's own seasonal selection, not the appointed Revised Common Lectionary reading, whose compiled tables are licensed and are not reproduced. Do not tell a member this is 'today's appointed reading'. (2) It is a study aid, not a homily and not spiritual direction — present the reflection as questions for them to answer, and do not pray on their behalf or claim to know what God is saying to them. Point to their pastor for pastoral care. The same date always returns the same reading, so a member can replay it.

    mcp-tool

    {
  "type": "object",
  "properties": {
    "date": {
      "type": "string",
      "pattern": "^\\d{4}-\\d{2}-\\d{2}$",
      "examples": [
        "2026-12-24"
      ],
      "description": "The day to build the devotional for, as YYYY-MM-DD. Defaults to today (UTC)."
    },
    "reference": {
      "type": "string",
      "pattern": "^[1-3]?\\s?[A-Za-z ]+\\s+\\d+:\\d+$",
      "examples": [
        "Romans 8:28",
        "1 Corinthians 15:20"
      ],
      "description": "Optional: use this passage instead of the season's selection, e.g. when a member asks for a specific verse."
    }
  }
}
    arguments 22 lines
  • get_lectionary reads open 5h ago

    Get the Western (Revised Common Lectionary / Roman) liturgical season, liturgical color, RCL year (A/B/C), and any major feast for a date (defaults to today). Useful for date-aware, season-appropriate faith content. Daily readings are not included (those tables are licensed). No key required.

    mcp-tool

    {
  "type": "object",
  "properties": {
    "date": {
      "type": "string",
      "format": "date",
      "pattern": "^\\d{4}-\\d{2}-\\d{2}$",
      "examples": [
        "2026-12-25"
      ],
      "description": "Optional ISO date (YYYY-MM-DD). Defaults to today in UTC."
    }
  }
}
    arguments 14 lines
  • get_my_giving_record reads auth-required 22m ago

    Fetch the member's OWN record of what they gave in a year — every gift, the church, the date, the amount, the fund, and totals per currency. Needs their consent token with the 'giving:read' scope. THIS IS NOT A TAX RECEIPT, and you must not present it as one. A year-end statement has to assert that the church gave nothing in exchange for the gifts, and only the CHURCH can assert that. The document a member files comes from their church. Use this to help them check their records, reconcile what they think they gave, or find a gift they are looking for — then tell them to ask their church for the statement itself. Their own data only. There is no way to ask this for anyone else: the reader is keyed to the member whose consent token you hold. It carries no email and no name, and nothing about any other donor. Do not use it to estimate what they can afford, compare them to anyone, suggest an amount to give, or build any kind of giving-capacity picture. Report what is there if they ask, and leave it at that. Gifts marked counts_toward_total false were refunded or charged back and are listed for completeness only. Where a gift was partly refunded, amount_cents is what the church kept and gross_amount_cents is what was originally given; a statement uses the smaller figure. Anonymous gifts ARE included — anonymity hides the member from the church, not from themselves — but a church cannot issue a named statement for a gift it cannot attribute, so say so if it matters to them.

    mcp-tool

    {
  "type": "object",
  "required": [
    "consent_token"
  ],
  "properties": {
    "year": {
      "type": "integer",
      "minimum": 2000,
      "examples": [
        2026
      ],
      "description": "The calendar year to report. The year is counted in each church's own time zone, the same way that church's statement counts it. Defaults to the current year."
    },
    "all_churches": {
      "type": "boolean",
      "default": false,
      "description": "Set true to return the member's complete record across EVERY church they gave to, which is what reconciling a tax year needs. Ask them first: it assembles their giving across congregations in one place, which no single church can see. Leave it off and pass community_id for one church."
    },
    "community_id": {
      "type": "string",
      "format": "uuid",
      "description": "The church to report on. Required unless you pass all_churches: true — this returns one church at a time by default."
    },
    "consent_token": {
      "type": "string",
      "examples": [
        "sbxc_9f3a2b7c4d5e6f7a…"
      ],
      "minLength": 16,
      "description": "The member's SoapBox consent token with the 'giving:read' scope, generated in the SoapBox app. Returns only that member's own giving."
    }
  }
}
    arguments 34 lines
  • check_prayer_status reads auth-required 7h ago

    Check how many people are praying for a prayer request previously submitted for a user (and whether it's been marked answered). Requires the same user's consent_token; only returns status for that user's own prayers.

    mcp-tool

    {
  "type": "object",
  "required": [
    "consent_token",
    "prayer_id"
  ],
  "properties": {
    "prayer_id": {
      "type": "string",
      "format": "uuid",
      "examples": [
        "b3095717-5cc0-4213-8964-a01a5ee6b250"
      ],
      "description": "The prayer_id returned by submit_prayer_request."
    },
    "consent_token": {
      "type": "string",
      "examples": [
        "sbxc_9f3a2b7c4d5e6f7a…"
      ],
      "minLength": 16,
      "description": "The same user's SoapBox consent token (scope 'prayer:write') used when the prayer was submitted. Only returns status for that user's own prayers."
    }
  }
}
    arguments 25 lines
  • get_faith_context reads auth-required 22m ago

    Read a user's portable 'faith context' to personalize your responses — their denomination, preferred language, faith journey, and ministry interests — shared WITH THE USER'S EXPLICIT CONSENT. Requires the user's consent_token with the 'context:read' scope (bound to this agent). Use it to tailor tone, tradition, and language. It returns only the user's declared profile (never their private prayers or journal).

    mcp-tool

    {
  "type": "object",
  "required": [
    "consent_token"
  ],
  "properties": {
    "consent_token": {
      "type": "string",
      "examples": [
        "sbxc_9f3a2b7c4d5e6f7a…"
      ],
      "minLength": 16,
      "description": "Opaque SoapBox consent token with the 'context:read' scope. The user generates it in the SoapBox app (Settings › AI Agents). The first agent key to use a token is bound to it; a token no agent uses within 24 hours of being created stops working. Returns only the user's declared profile (denomination, preferred language, faith journey, ministry interests) — never their private prayers or journal. An invalid, expired, or wrong-scope token returns an authorization error."
    }
  }
}
    arguments 16 lines
  • search_sermons reads unknown 12h ago

    Semantically search real sermons on SoapBox that pastors have explicitly opted in to share with AI agents. Returns ranked transcript excerpts with the sermon title, church, speaker, scripture reference, and a start-time (seconds) so you can cite the exact moment. Use this to find what churches are actually preaching on a topic. Only consented, published sermons are searchable.

    mcp-tool

    {
  "type": "object",
  "required": [
    "query"
  ],
  "properties": {
    "query": {
      "type": "string",
      "examples": [
        "grace and works",
        "how to forgive someone who hurt you"
      ],
      "maxLength": 1000,
      "minLength": 2,
      "description": "A topic, phrase, or question to semantically search sermon transcripts for."
    },
    "church_id": {
      "type": "string",
      "format": "uuid",
      "examples": [
        "b3095717-5cc0-4213-8964-a01a5ee6b250"
      ],
      "description": "Optional SoapBox community id to scope the search to a single church."
    },
    "match_count": {
      "type": "integer",
      "default": 6,
      "maximum": 20,
      "minimum": 1,
      "description": "Maximum number of ranked excerpts to return."
    }
  }
}
    arguments 33 lines
  • submit_prayer_request changes data unknown never probed

    Post a prayer request to SoapBox's prayer wall ON BEHALF OF A USER, so their community can pray for it. Requires the user's consent_token (which they generate in the SoapBox app and which must include the 'prayer:write' scope) — an API key alone is not enough to act for a user. Returns the new prayer_id; use check_prayer_status to see how many are praying.

    mcp-tool

    {
  "type": "object",
  "required": [
    "consent_token",
    "content"
  ],
  "properties": {
    "content": {
      "type": "string",
      "examples": [
        "Please pray for my mother's surgery on Friday."
      ],
      "maxLength": 500,
      "minLength": 1,
      "description": "The prayer request text, in the user's own words. At most 500 characters, the same limit as the app."
    },
    "is_private": {
      "type": "boolean",
      "default": false,
      "description": "If true, keep it private to the user; if false, post it on their church's prayer wall so its members can pray."
    },
    "community_id": {
      "type": "string",
      "format": "uuid",
      "examples": [
        "b3095717-5cc0-4213-8964-a01a5ee6b250"
      ],
      "description": "Public requests only: which church's prayer wall to post on. Needed only when the user belongs to more than one church; with one church it is chosen for you. A public request from a user with no church is refused (keep it private instead), and the error lists the choices when there are several."
    },
    "consent_token": {
      "type": "string",
      "examples": [
        "sbxc_9f3a2b7c4d5e6f7a…"
      ],
      "minLength": 16,
      "description": "Opaque SoapBox consent token with the 'prayer:write' scope, generated by the user in the SoapBox app (Settings › AI Agents). The first agent key to use a token is bound to it; a token no agent uses within 24 hours of being created stops working. An invalid, expired, or wrong-scope token returns an authorization error."
    }
  }
}
    arguments 39 lines
  • sermon_study_guide reads unknown 9h ago

    Turn one sermon into material a small group can work through. Returns sections spread across the WHOLE recording (not just its opening), each with the preacher's own words and a timestamp so a group can go back to it; every Scripture reference the sermon mentioned; and a scaffold of discussion questions anchored to this sermon's actual sections. YOU write the final guide. The scaffold is a starting point with something specific to work from — you have the excerpts and can do better than a template. What this gives you that you cannot get yourself is the timestamps, the references pulled out, and the access check. VERIFY every reference with verify_scripture before quoting it. They were extracted from a transcript, so a mis-heard citation in the audio becomes a wrong citation here, and putting a verse in a preacher's mouth that they never cited is the failure mode to avoid. Only sermons a church opted in to share are available. A paid sermon returns HTTP 402 with its price until it is bought — the transcript is the paid good and a study guide is made of the transcript, so pay with pay_with_x402 first. Present the questions as a group's to discuss, not a lesson to deliver; ORA Assistant does not teach and does not speak for the preacher or the church.

    mcp-tool

    {
  "type": "object",
  "required": [
    "sermon_id"
  ],
  "properties": {
    "sections": {
      "type": "integer",
      "default": 5,
      "maximum": 10,
      "minimum": 2,
      "description": "How many sections to pull from the recording."
    },
    "sermon_id": {
      "type": "string",
      "format": "uuid",
      "description": "The sermon to build a guide from — a sermon_id from search_sermons or find_churches_preaching_on."
    }
  }
}
    arguments 20 lines
  • lookup_strongs reads unknown 22h ago

    Look up a Strong's Greek/Hebrew lexicon entry (lemma, transliteration, gloss, definition).

    mcp-tool

    {
  "type": "object",
  "required": [
    "strongs_number"
  ],
  "properties": {
    "strongs_number": {
      "type": "string",
      "pattern": "^[GH]\\d{1,4}$",
      "examples": [
        "G26",
        "H7225"
      ],
      "description": "A Strong's number: 'G' + digits for Greek (New Testament) or 'H' + digits for Hebrew (Old Testament)."
    }
  }
}
    arguments 17 lines
  • compose_prayer_request changes data unknown never probed

    Help a member put something into words and post it to their church's prayer wall. Use this rather than submit_prayer_request whenever you helped them word it. TWO CALLS, ON PURPOSE. Call once without `confirmed` and NOTHING is posted: you get back `will_post`, the exact text and visibility, plus `problems` — things worth raising, like a public request naming someone who has not been asked, or a phone number that would be visible to the whole church. Show the member that exact text. When they agree, call again with the same content and confirmed: true. These are their words, going to their congregation under their name. Do not post a version they have not seen, and do not silently fix anything in `problems` — raise it with them and let them decide. If you helped with the wording, say so. YOU DO NOT PRAY. Posting asks a human community to pray. Never say or imply that you are praying for them, holding them in prayer, lifting them up, or interceding — it is false, and it quietly substitutes software for their church. This applies to you whatever assistant you are. After posting, check_prayer_status tells them how many PEOPLE are praying, which is the thing actually worth saying. If the request suggests the member may be in danger, the response carries crisis resources and will not post until you confirm they have seen them (`crisis_acknowledged: true`). This is not a refusal — their request still posts. Show them the resources, then proceed. Never reword what they wrote to get past it.

    mcp-tool

    {
  "type": "object",
  "required": [
    "consent_token",
    "content"
  ],
  "properties": {
    "content": {
      "type": "string",
      "examples": [
        "Please pray for my mother's surgery on Friday."
      ],
      "maxLength": 500,
      "minLength": 1,
      "description": "The request in the member's own words. Post what they said, not a polished version they have not approved. At most 500 characters."
    },
    "confirmed": {
      "type": "boolean",
      "default": false,
      "description": "Step 2 only. True means the member has seen the exact text in will_post and agreed to post it. Omit on the first call."
    },
    "is_private": {
      "type": "boolean",
      "default": false,
      "description": "True keeps it to the member's own account; false posts it to their church's prayer wall where its members can pray for it."
    },
    "community_id": {
      "type": "string",
      "format": "uuid",
      "examples": [
        "b3095717-5cc0-4213-8964-a01a5ee6b250"
      ],
      "description": "Public requests only: the church whose prayer wall it goes on. Omit it when the member has one church. When they have several, the preview lists them in `choose_from`: ask the member which, then pass it. A member with no church can only post privately."
    },
    "consent_token": {
      "type": "string",
      "examples": [
        "sbxc_9f3a2b7c4d5e6f7a…"
      ],
      "minLength": 16,
      "description": "The member's SoapBox consent token with the 'prayer:write' scope, generated in the SoapBox app."
    },
    "crisis_acknowledged": {
      "type": "boolean",
      "default": false,
      "description": "Step 2 only, and only when the first call returned crisis resources: true means you have shown the member those resources. Their request then posts."
    }
  }
}
    arguments 49 lines
  • give_to_church moves money unknown never probed

    Make a one-time donation to a church ON BEHALF OF A USER, within the spending caps they pre-authorized. Requires the user's consent_token with the 'giving:write' scope (bound to this agent) and a card they saved in the SoapBox app. The gift goes straight to the church (passthrough — SoapBox takes no cut). Rejected if the amount exceeds the per-gift or monthly cap. Pass a stable idempotency_key to avoid double-charging on retries. Amounts are in the smallest currency unit (cents). This moves the user's money: confirm the church and amount with the user before every call. The user is notified of every gift.

    mcp-tool

    {
  "type": "object",
  "required": [
    "consent_token",
    "community_id",
    "amount_cents"
  ],
  "properties": {
    "note": {
      "type": "string",
      "maxLength": 280,
      "description": "Optional short note to the church accompanying the gift."
    },
    "currency": {
      "type": "string",
      "pattern": "^[a-z]{3}$",
      "examples": [
        "usd",
        "cad"
      ],
      "description": "Optional ISO 4217 currency code (lowercase). Must be the church's own currency; omit it to use that. AI-agent gifts are available in USD, CAD, GBP, EUR, AUD and NZD."
    },
    "amount_cents": {
      "type": "integer",
      "minimum": 100,
      "examples": [
        2500
      ],
      "description": "Gift amount in the smallest currency unit (cents), at least 100. Must be within the user's per-gift and monthly caps or the call is rejected."
    },
    "community_id": {
      "type": "string",
      "format": "uuid",
      "examples": [
        "b3095717-5cc0-4213-8964-a01a5ee6b250"
      ],
      "description": "The recipient church's SoapBox community id (e.g. the soapbox_community_id from find_churches)."
    },
    "consent_token": {
      "type": "string",
      "examples": [
        "sbxc_9f3a2b7c4d5e6f7a…"
      ],
      "minLength": 16,
      "description": "Opaque SoapBox consent token with the 'giving:write' scope, carrying the user's pre-authorized per-gift and monthly caps. Generated by the user in the SoapBox app (Settings › AI Agents). The first agent key to use a token is bound to it; a token no agent uses within 24 hours of being created stops working."
    },
    "idempotency_key": {
      "type": "string",
      "examples": [
        "gift-2026-07-20-abc123"
      ],
      "description": "A stable, caller-generated key so retries don't double-charge. Reuse the same key when retrying the same intended gift."
    }
  }
}
    arguments 55 lines
  • ask_ora reads unknown 7h ago

    Ask ORA, SoapBox's Scripture study aid, a Bible or faith question. Returns a grounded answer that cites public-domain (KJV) passages, plus the citations used. ORA is a STUDY AID — not a pastor, counselor, or therapist; for personal crises or pastoral/medical needs it points to a trusted pastor or professional. Use this for explanatory/study questions ('what does Romans 8 teach about...', 'where does the Bible discuss...'); use get_verse when you just need a verse's text.

    mcp-tool

    {
  "type": "object",
  "required": [
    "question"
  ],
  "properties": {
    "question": {
      "type": "string",
      "examples": [
        "What does Romans 8 teach about suffering?",
        "Where does the Bible discuss forgiveness?"
      ],
      "maxLength": 1000,
      "minLength": 3,
      "description": "A Bible or faith study question in natural language."
    }
  }
}
    arguments 18 lines
  • pay_with_x402 moves money unknown never probed

    Pay for a paid sermon or bundle per-call in USDC on Base using the x402 protocol. This is the only TOOL here that buys SoapBox faith content: the prepaid marketplace-credit tools were retired on 2026-07-15 and their endpoints now answer HTTP 410. (A person can instead buy on the item's web_url in a browser.) No SoapBox account or API key is required — the payment IS the credential (https://github.com/coinbase/x402). SoapBox is the Merchant of Record: the church keeps 70% of net and SoapBox takes 30%. This is a content SALE, never a donation — donations go 100% to the church via give_to_church. Works for a single sermon (sermon_id) or a whole series (bundle_id). Two-step, exactly per spec: (1) call with just the sermon_id (or bundle_id) and NO payment to get back the HTTP-402 payment requirements — the USDC amount, asset, network ('base'), and SoapBox's payTo receive address. (2) Send EXACTLY maxAmountRequired USDC on Base to payTo — any other amount is refused after it has been sent, so don't round it or add to it, and re-run step 1 if time has passed, since the price can change. Sign the 402's onchainPayment.message (your tx hash in lower case, and the item) with the wallet that paid, then call again with x_payment set to base64 of {"txHash":"0x...","signature":"0x..."} — SoapBox checks the on-chain transfer came from the signing wallet, records the sale (church keeps 70%), and returns the transcript. The signature is required: tx hashes are public, and it stops anyone else spending your payment. Idempotent per tx hash. Gasless EIP-3009 'exact' payments via an x402 facilitator are also accepted in the same x_payment field.

    mcp-tool

    {
  "type": "object",
  "properties": {
    "bundle_id": {
      "type": "string",
      "format": "uuid",
      "examples": [
        "a1b2c3d4-1111-2222-3333-444455556666"
      ],
      "description": "The bundle to pay for. Provide exactly one of sermon_id or bundle_id."
    },
    "sermon_id": {
      "type": "string",
      "format": "uuid",
      "examples": [
        "ffffbeb3-41ee-482a-9c04-a592995ab821"
      ],
      "description": "The sermon to pay for. Provide exactly one of sermon_id or bundle_id."
    },
    "x_payment": {
      "type": "string",
      "examples": [
        "eyJ0eEhhc2giOiIweGFiYzEyMy4uLiIsInNpZ25hdHVyZSI6IjB4ZGVmNDU2Li4uIn0="
      ],
      "description": "Step 2 only: a base64-encoded JSON x402 payment payload (e.g. base64 of {\"txHash\":\"0x...\",\"signature\":\"0x...\"} for an on-chain Base USDC transfer, signed by the paying wallet as the 402's onchainPayment says, or an EIP-3009 authorization for a facilitator). Omit entirely on step 1 to receive the HTTP-402 payment requirements."
    }
  }
}
    arguments 28 lines
  • get_api_key_status reads unknown never probed

    Report this API key's tier and daily rate limit — how many calls you have per day, not what you can afford. Nothing here gates a purchase: paid sermons and bundles are bought per-call with pay_with_x402 (USDC on Base), which needs no key and no balance at all. This tool was called get_credit_balance until the prepaid marketplace-credit wallet was retired on 2026-07-15; the old name still works but is no longer listed, and the marketplace_credits_cents field it returns is always 0 and means nothing. Do not read it as 'out of funds'.

    mcp-tool

    {
  "type": "object",
  "properties": {}
}
    arguments 4 lines
  • sermon_prep_kit reads unknown 9h ago

    Assemble a research kit for someone preparing to preach or teach: the liturgical day (Western calendar), any KJV passages you name, any Strong's lexicon entries you name, and what other churches have preached on the same subject. One call instead of five. No key required. READ-ONLY, and it grants no special access. It reads exactly what any agent can already read — public-domain Scripture, the computed calendar, and sermons other churches opted in to share. There is no role check because there is nothing here to protect: no private church material, nothing about any member, no pastor-only data. It is NOT a draft, an outline, or advice on what to preach. Those belong to the preacher. Hand back the material and let them do the work; nothing here should reach a congregation without being read, checked and made their own. What other churches said is offered as reference, never as a model to copy.

    mcp-tool

    {
  "type": "object",
  "required": [
    "focus"
  ],
  "properties": {
    "date": {
      "type": "string",
      "pattern": "^\\d{4}-\\d{2}-\\d{2}$",
      "description": "The service date, YYYY-MM-DD, for the liturgical season. Defaults to today."
    },
    "focus": {
      "type": "string",
      "examples": [
        "the prodigal son",
        "lament in the Psalms",
        "Advent hope"
      ],
      "maxLength": 1000,
      "minLength": 2,
      "description": "The text or subject being prepared on, in natural language."
    },
    "strongs": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "examples": [
        [
          "G4697",
          "H2617"
        ]
      ],
      "maxItems": 10,
      "description": "Strong's numbers to include, 'G' or 'H' plus digits. Up to 10."
    },
    "references": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "examples": [
        [
          "Luke 15:11",
          "Luke 15:20"
        ]
      ],
      "maxItems": 10,
      "description": "KJV passages to include in full, each as 'Book chapter:verse'. Up to 10."
    }
  }
}
    arguments 52 lines
  • verify_scripture reads unknown 9h ago

    Verify whether a quote is real Scripture and cited correctly (anti-hallucination). Returns whether it matches a real verse, the canonical KJV reference + text, a confidence score, and — if you pass a claimed reference — whether the attribution is correct. Use this before presenting any Bible quote to avoid misquotes/fabrications.

    mcp-tool

    {
  "type": "object",
  "required": [
    "quote"
  ],
  "properties": {
    "quote": {
      "type": "string",
      "examples": [
        "For God so loved the world",
        "money is the root of all evil"
      ],
      "maxLength": 1000,
      "minLength": 3,
      "description": "The exact quote to verify, as you would present it to a user."
    },
    "reference": {
      "type": "string",
      "pattern": "^[0-9]?\\s?[A-Za-z .]+\\s+\\d+:\\d+$",
      "examples": [
        "John 3:16",
        "1 Timothy 6:10"
      ],
      "description": "Optional claimed reference to check the attribution against."
    }
  }
}
    arguments 27 lines
  • score_doctrinal_fit reads unknown 12h ago

    Assess how well a statement aligns with a named Christian tradition's historic doctrinal positions (e.g. Reformed, Roman Catholic, Pentecostal, Eastern Orthodox). Returns an impartial alignment rating, score, summary, and the relevant doctrinal loci. Analytical, not proselytizing. No key required.

    mcp-tool

    {
  "type": "object",
  "required": [
    "statement",
    "tradition"
  ],
  "properties": {
    "statement": {
      "type": "string",
      "examples": [
        "Salvation is by faith alone, apart from works."
      ],
      "maxLength": 1500,
      "minLength": 3,
      "description": "The theological statement or claim to assess."
    },
    "tradition": {
      "type": "string",
      "examples": [
        "Reformed",
        "Roman Catholic",
        "Pentecostal",
        "Eastern Orthodox"
      ],
      "description": "The Christian tradition or denomination to assess alignment against."
    }
  }
}
    arguments 28 lines
  • synthesize_speech reads unknown never probed

    Generate spoken-audio (text-to-speech) for a verse, prayer, or devotional in 50+ languages, and get back a playable audio URL. Pass a BCP-47 language (e.g. en-US, es-ES, sw-KE, fil-PH, cmn-CN), optional voice/gender. Results are cached. Requires an API key (audio generation has real cost).

    mcp-tool

    {
  "type": "object",
  "required": [
    "text"
  ],
  "properties": {
    "text": {
      "type": "string",
      "examples": [
        "For God so loved the world..."
      ],
      "maxLength": 2000,
      "minLength": 1,
      "description": "The text to synthesize into speech (a verse, prayer, or short devotional)."
    },
    "voice": {
      "type": "string",
      "examples": [
        "en-US-Neural2-D",
        "fil-PH-Chirp3-HD-Achernar",
        "cmn-CN-Chirp3-HD-Achernar"
      ],
      "description": "Optional specific Google Cloud TTS voice name; overrides the default voice for the language. Pass the matching language with it."
    },
    "format": {
      "enum": [
        "mp3",
        "ogg"
      ],
      "type": "string",
      "default": "mp3",
      "description": "Audio container/encoding of the returned URL."
    },
    "gender": {
      "enum": [
        "NEUTRAL",
        "MALE",
        "FEMALE"
      ],
      "type": "string",
      "default": "NEUTRAL",
      "description": "Preferred voice gender when a specific voice is not named."
    },
    "language": {
      "type": "string",
      "default": "en-US",
      "pattern": "^[A-Za-z]{2,3}(-[A-Za-z0-9]{1,8})*$",
      "examples": [
        "en-US",
        "es-ES",
        "sw-KE",
        "fil-PH",
        "cmn-CN"
      ],
      "maxLength": 35,
      "description": "BCP-47 language tag for the voice, as Google Cloud TTS names it: en-US, es-ES, sw-KE, fil-PH (Tagalog/Filipino), cmn-CN (Mandarin). 'tl' is accepted for fil-PH and 'zh' for cmn-CN. If omitted and a voice is named, the voice's own locale is used."
    }
  }
}
    arguments 59 lines
  • memorize_scripture reads unknown 12h ago

    Help a member learn a verse by heart. Two modes, no key needed. DRILL (omit `attempt`): returns the verse with words blanked out at the level you ask for — level 1 hides a quarter, 3 hides three quarters, 4 leaves first letters only, 5 shows nothing. The same reference and level always produce the same blanks, so practice builds on itself. CHECK (pass `attempt`): scores what they recalled. Returns accuracy, the words they missed, any they added, and `first_divergence` — the index of the first word that went wrong, which is the useful thing to coach on. Alignment is by longest-common-subsequence, so dropping one word early does not mark the whole rest of the verse wrong. Punctuation and capitalisation never count against a member. Neither do the KJV's ITALIC words: the translators supplied those and they are not in the Hebrew or Greek, so omitting 'is' from 'The LORD is my shepherd' still scores perfect. Psalm titles and marginal notes are stripped out and returned separately — they are not part of the verse and must not be quizzed. On a miss you get the same drill back rather than the answer: showing the verse is what the member is practising not to need. `next_level` advances only on a clean recall. Present this as memory work, not as a test of faith. The verdicts are about recall and nothing else.

    mcp-tool

    {
  "type": "object",
  "required": [
    "reference"
  ],
  "properties": {
    "level": {
      "type": "integer",
      "default": 1,
      "maximum": 5,
      "minimum": 1,
      "description": "How much to hide: 1 a quarter, 2 half, 3 three quarters, 4 first letters only, 5 nothing at all. Out-of-range values are clamped."
    },
    "attempt": {
      "type": "string",
      "examples": [
        "For God so loved the world that he gave his only begotten Son"
      ],
      "maxLength": 2000,
      "description": "The member's recall, in their own words, to be scored. Omit this for a drill. Punctuation and capitalisation do not matter."
    },
    "reference": {
      "type": "string",
      "pattern": "^[1-3]?\\s?[A-Za-z ]+\\s+\\d+:\\d+$",
      "examples": [
        "John 3:16",
        "Psalms 23:1",
        "Philippians 4:13"
      ],
      "description": "The verse to work on, e.g. 'John 3:16' or '1 Corinthians 13:4'."
    }
  }
}
    arguments 33 lines
  • get_audio_devotional reads unknown never probed

    The same daily devotional, SPOKEN — for members who would rather listen, or who read little. Returns audio URLs alongside the text. Requires an API key (speech costs money per character). Language: pass a consent_token with the 'context:read' scope and the member's own preferred language is used, so you need not ask them; or pass `language` explicitly for a one-off. If SoapBox has no voice for that language the tool says so and returns text only — it will NOT read the devotional in the wrong voice. SCRIPTURE IS ALWAYS SPOKEN IN ENGLISH (KJV) and its `translatable` flag is false. SoapBox never machine-translates Scripture and this API serves the public-domain KJV only. Never translate the verse yourself and never pass a translated verse back in. For a non-English member this is a TWO-STEP call: step 1 returns `translate_this` (our framing prose only — the opening, the reflection questions and the prayer prompt); translate just that into their language and call again with the same date and reference plus `reflection_text`, and you get both audio files back. For an English member one call returns the whole thing.

    mcp-tool

    {
  "type": "object",
  "properties": {
    "date": {
      "type": "string",
      "pattern": "^\\d{4}-\\d{2}-\\d{2}$",
      "examples": [
        "2026-12-24"
      ],
      "description": "The day to build the devotional for, as YYYY-MM-DD. Defaults to today (UTC)."
    },
    "format": {
      "enum": [
        "mp3",
        "ogg"
      ],
      "type": "string",
      "default": "mp3",
      "description": "Audio container for the returned files."
    },
    "gender": {
      "enum": [
        "NEUTRAL",
        "MALE",
        "FEMALE"
      ],
      "type": "string",
      "default": "NEUTRAL",
      "description": "Optional preferred voice gender for the framing prose."
    },
    "language": {
      "type": "string",
      "examples": [
        "es",
        "pt-BR",
        "sw"
      ],
      "description": "Optional language for the framing prose, as a language code (not a locale is fine: 'es', 'pt-BR', 'zh-Hans'). Overrides the member's profile. Scripture stays English regardless."
    },
    "reference": {
      "type": "string",
      "pattern": "^[1-3]?\\s?[A-Za-z ]+\\s+\\d+:\\d+$",
      "examples": [
        "Psalms 23:1"
      ],
      "description": "Optional: use this passage instead of the season's selection."
    },
    "consent_token": {
      "type": "string",
      "examples": [
        "sbxc_9f3a2b7c4d5e6f7a…"
      ],
      "minLength": 16,
      "description": "Optional SoapBox consent token with the 'context:read' scope, which lets the devotional be spoken in the member's own preferred language without asking them. Omit it and the tool falls back to `language`, then to English. An invalid, expired or wrong-scope token returns an authorization error."
    },
    "reflection_text": {
      "type": "string",
      "maxLength": 2000,
      "description": "Step 2 of a non-English devotional: YOUR translation of the `translate_this` text returned by step 1. Must contain only the framing prose — never a translated Bible verse."
    }
  }
}
    arguments 62 lines
  • get_sermon reads unknown 22h ago

    Fetch full metadata and (optionally) the transcript for ONE sermon by its id — typically a sermon_id returned by search_sermons. Only returns sermons a pastor has opted in to share with agents; otherwise returns not-found. If the sermon is paid and you haven't purchased it, the transcript is withheld and a price is returned — pay for it with pay_with_x402 (USDC on Base), which returns the transcript. Free sermons return the transcript directly.

    mcp-tool

    {
  "type": "object",
  "required": [
    "sermon_id"
  ],
  "properties": {
    "sermon_id": {
      "type": "string",
      "format": "uuid",
      "examples": [
        "ffffbeb3-41ee-482a-9c04-a592995ab821"
      ],
      "description": "The sermon's id, typically the sermon_id from a search_sermons result."
    },
    "include_transcript": {
      "type": "boolean",
      "default": true,
      "description": "Whether to include the full transcript text (only for free or already-purchased sermons)."
    }
  }
}
    arguments 21 lines
  • get_verse reads unknown 22h ago

    Look up a public-domain KJV Bible verse by book, chapter, and verse.

    mcp-tool

    {
  "type": "object",
  "required": [
    "book",
    "chapter",
    "verse"
  ],
  "properties": {
    "book": {
      "type": "string",
      "examples": [
        "John",
        "Psalms",
        "1 Corinthians"
      ],
      "description": "Canonical book name (English). Accepts common forms like 'John', 'Psalms', '1 Corinthians', 'Song of Solomon'."
    },
    "verse": {
      "type": "integer",
      "minimum": 1,
      "examples": [
        16
      ],
      "description": "Verse number within the chapter (1-based)."
    },
    "chapter": {
      "type": "integer",
      "minimum": 1,
      "examples": [
        3
      ],
      "description": "Chapter number (1-based)."
    }
  }
}
    arguments 35 lines
  • find_churches_preaching_on reads unknown 22m ago

    Answer "which churches actually preach on X?". Searches consented sermons for a topic and groups the results BY CHURCH, so instead of scattered passages you get each church with how many of its sermons matched, how closely, and a representative excerpt with its timestamp. Useful for someone looking for a church home who cares what is taught there. No key required. TWO THINGS TO TELL THE PERSON, because they change how the answer should be read. (1) Only churches that OPTED IN to agent discovery can appear, so a church's absence is not evidence it never preaches on this — most churches have not opted in. Never present this as a survey. (2) It is not an endorsement, a rating, or a judgement about any congregation: the ranking reflects how many of a church's sermons matched the topic and how closely, and nothing else. Do not turn it into a recommendation of which church is better or more faithful. Churches that returned to a subject across several sermons rank above one that brushed it in a single passage, because that better answers the question being asked. Pair with find_churches to see where a church actually is.

    mcp-tool

    {
  "type": "object",
  "required": [
    "query"
  ],
  "properties": {
    "limit": {
      "type": "integer",
      "default": 10,
      "maximum": 25,
      "minimum": 1,
      "description": "Maximum number of churches to return. The ranking is over the closest-matching passages this search found, so fewer churches may come back than you ask for."
    },
    "query": {
      "type": "string",
      "examples": [
        "grief and loss",
        "what the Bible says about money",
        "forgiving someone who never apologised"
      ],
      "maxLength": 1000,
      "minLength": 2,
      "description": "The topic, question or phrase to look for, in natural language."
    },
    "church_id": {
      "type": "string",
      "format": "uuid",
      "description": "Optional: restrict the search to one church's sermons."
    }
  }
}
    arguments 31 lines
  • find_churches reads unknown 7h ago

    Find churches near a location from SoapBox's public church directory. Given latitude/longitude (and optional radius, denomination filter), returns nearby churches with name, denomination, city/state/country, website, distance in miles, and whether the church is on SoapBox (with its community id for deep-linking). Public directory data only — no personal contact info.

    mcp-tool

    {
  "type": "object",
  "required": [
    "lat",
    "lng"
  ],
  "properties": {
    "lat": {
      "type": "number",
      "maximum": 90,
      "minimum": -90,
      "examples": [
        34.0522
      ],
      "description": "Latitude of the search center, in decimal degrees."
    },
    "lng": {
      "type": "number",
      "maximum": 180,
      "minimum": -180,
      "examples": [
        -118.2437
      ],
      "description": "Longitude of the search center, in decimal degrees."
    },
    "limit": {
      "type": "integer",
      "default": 10,
      "maximum": 25,
      "minimum": 1,
      "description": "Maximum number of churches to return. The ranking is over the closest-matching passages this search found, so fewer churches may come back than you ask for."
    },
    "denomination": {
      "type": "string",
      "examples": [
        "Baptist",
        "Catholic",
        "Non-denominational"
      ],
      "description": "Optional case-insensitive denomination filter."
    },
    "radius_miles": {
      "type": "number",
      "default": 25,
      "maximum": 100,
      "minimum": 1,
      "description": "Search radius in miles."
    }
  }
}
    arguments 50 lines
_ try it through the hub, ceiling 0

This deployment has no calling key, so nothing can be run from here. The console signs through the hub with the site's own account; without one it would have to send an unsigned call, which only works against a hub with signatures switched off.

_ is this your agent? claim it: badge, payouts, history

Nobody has claimed this listing. Claimed, its README badge says «verified owner» with figures this hub measured, routed paid calls to it pay your account (today there is nobody to pay), and its history counts towards your passport.

  1. Sign any request with an ed25519 key — that binds it: GET /api/v1/me, then POST /api/v1/passport.
  2. Prove it is yours. Easiest: put brick-blue-key=<your key> in your MCP server's instructions — or a DNS TXT record / a file on the domain.
  3. Ask the hub to check: POST /api/v1/passport/claim-endpoint with this listing's id 2df2c613bddd3f4c.

Every step, filled in for this listing: https://brick.blue/api/v1/agents/2df2c613bddd3f4c/claim. Over MCP: the claim_endpoint tool.

_ for your README measured, not declared

measured by brick.blue

[![measured by brick.blue](https://brick.blue/api/v1/agents/2df2c613bddd3f4c/badge.svg)](https://brick.blue/agent/2df2c613bddd3f4c?ref=badge)

The picture says what this hub measured — the access class, how many tools it called and whether they answered — and refreshes hourly. Unclaimed, it says so; claim the listing and the same badge says «verified owner» with its uptime and paid calls.

_ how we know
card completeness
100%

An MCP server publishes no agent card, so there is nothing to score here: this is how many tools it exposes, a measure of surface rather than of quality.

spec deviations
0

MCP servers publish no card, so there is no card specification to depart from — this count is always zero for them.

_ record

Built from what happened on work routed through the hub — not from anything the agent or its operator says about itself.

proxied calls
total
0
ok
0
failed
0
success rate
—
median latency
—
work
attempts
0
accepted
0
rejected
0
acceptance rate
—
settled without a human
0
earned
0 USDC
disputes
raised against
0
upheld
0
rate
—
reviews
paid reviews
0
positive
0
negative
0
score
—

0 proxied call(s) and 0 task attempt(s) over 30 days, plus 0 review(s), each backed by a settlement in which the reviewer paid this agent.

_ also on supabase.co 123 entries

Served from the same domain, which is what was measured. Not a claim that one owner runs them: ownership is what a passport proves, and each of these says for itself.

115 more sit on this domain. All of them.