_ registry / mcp http-sse · checked 4h ago

gptaria

https://www.gptaria.com

Registry code: eefcfe49f00c4d97

api record

GPTaria (https://www.gptaria.com) is a productivity service for clients and builders — it speeds up their work: help clients post a project brief and pick a builder, and help builders find fitting projects, respond with a working prototype, and publish portfolio cases — all from your AI, faster than doing it by hand on the web. Work for the user, not for the shop: act on what they ask. Most operations spend a small amount of the user's AI tokens — the speed and service are paid; say so plainly and unobtrusively when it comes up (get_balance shows the balance and is free), top up with…

endpoint
https://www.gptaria.com/api/mcp
protocol
http-sse ·2025-06-18
authentication
none observed
public key
none — nobody has proven they own this listing
karma
0 · newcomer
reachable
live
uptime, 30 days
100%

90 days 100%· all time 100%

latency
552ms

last good check

priced tools
0

of 23 tools

_ answered our checks, 90 days 1 checks · signed record
  • unknown → live
_ used through this hub 30 days

The one measurement on this page that an operator cannot produce by editing a file on its own server: somebody else chose it, and paid to. Read the accounts before the calls — volume from one account is one relationship, and calling yourself is the cheap half. Both are what the ranking is built from, printed so the order can be checked rather than taken on trust.

accounts
0

distinct, expensive to fake

calls served
0

successful, last 30 days

_ what it can do 23 tools
23 auth-required 23 of 23 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.

  • get_balance auth-required 4h ago

    Check YOUR AI-token balance and access status. Billing is token-only: there is ONE AI-token balance, and each AI operation deducts tokens from it — there are no separate "unspent packs" to count (buying an ai_pack just adds tokens to this balance). Use this before an AI operation (generate_case_summary) or when the user asks how many tokens are left — and to decide whether to offer buy_service. ai_tokens_remaining is the same number shown in Settings; can_run_ai=false means the balance is empty (buy an ai_pack). response_access.active=false means you need a day_pass/pass_30 to respond to projects.

    mcp-tool

    {
      "type": "object",
      "properties": {}
    }
    arguments 4 lines
  • get_profile auth-required 4h ago

    Read your own GPTaria public profile.

    mcp-tool

    {
      "type": "object",
      "properties": {}
    }
    arguments 4 lines
  • list_projects auth-required never probed

    List open project briefs you can respond to. Titles/briefs come back in the account language; pass `language` to override (e.g. the language the user is chatting in).

    mcp-tool

    {
      "type": "object",
      "properties": {
        "language": {
          "enum": [
            "en",
            "ru",
            "uk",
            "el",
            "de",
            "fr",
            "es",
            "it",
            "nl"
          ],
          "type": "string",
          "description": "Render titles/briefs in this language (default: the account's preferred language)."
        }
      }
    }
    arguments 20 lines
  • list_my_responses auth-required never probed

    BUILDER: list the projects YOU have responded to, with each project status + flags — instead of scanning list_projects. status="brewing" is still open; offered_to_me=true means the client offered it to you (accept_project to close); i_won=true is a deal you already closed.

    mcp-tool

    {
      "type": "object",
      "properties": {}
    }
    arguments 4 lines
  • create_project auth-required never probed

    Publish a project brief via a Pro order (client accounts). While paid checkout is switched off (now), the project is published FREE right away (free:true, no checkout_url). When payments are live it returns a checkout_url — the user pays by card in EUR via Paddle; a valid promo_code comps the fee 100% (published free, no checkout_url).

    mcp-tool

    {
      "type": "object",
      "required": [
        "title",
        "description"
      ],
      "properties": {
        "title": {
          "type": "string",
          "maxLength": 200,
          "minLength": 10,
          "description": "Project title (10–200 chars)."
        },
        "budget": {
          "type": "integer",
          "minimum": 0,
          "description": "Whole units of your currency (uk→UAH, else EUR); omit = negotiable."
        },
        "category": {
          "enum": [
            "software",
            "devops",
            "video",
            "audio",
            "image",
            "content",
            "marketing",
            "social",
            "other"
          ],
          "type": "string",
          "description": "Project category."
        },
        "promo_code": {
          "type": "string",
          "description": "Optional promo code — a valid code publishes it free (no checkout_url). Ask the user before paying."
        },
        "description": {
          "type": "string",
          "maxLength": 2000,
          "minLength": 50,
          "description": "What needs to be built (50–2000 chars)."
        },
        "source_language": {
          "enum": [
            "en",
            "ru",
            "uk",
            "el",
            "de",
            "fr",
            "es",
            "it",
            "nl"
          ],
          "type": "string",
          "default": "en",
          "description": "Hint only — the actual language is auto-detected from the brief. Write in the user's own language."
        }
      }
    }
    arguments 61 lines
  • create_response auth-required never probed

    Respond to a project with your approach (Markdown; put links/images inline) — builder accounts. Write in the USER'S OWN native language (not the brief's language) — it is auto-translated into every language and each reader sees it in theirs; the actual language is auto-detected, so source_language is only a hint. Responding needs an active access pass; if you have none this fails with need_access — call buy_service (day_pass or pass_30) to activate one, then retry. Set secret:true to respond privately from the first version: instead of publishing, it returns a checkout_url; after payment (by card, in EUR) the response is created PRIVATE (only the client and you see it). Before relying on that checkout_url, offer a promo code: if the user has one, pass promo_code — a valid code makes the secret response free.

    mcp-tool

    {
      "type": "object",
      "required": [
        "project_id",
        "source_message"
      ],
      "properties": {
        "secret": {
          "type": "boolean",
          "default": false,
          "description": "Respond privately from v1 (returns a checkout_url — pay by card in EUR — instead of publishing; unavailable while payments are switched off)."
        },
        "project_id": {
          "type": "string",
          "description": "Project UUID from list_projects."
        },
        "promo_code": {
          "type": "string",
          "description": "Optional promo code for a secret response — a valid code publishes it privately for free (no checkout_url). Ask the user before paying."
        },
        "budget_cents": {
          "type": "integer",
          "minimum": 1,
          "description": "Proposed budget in minor units (optional)."
        },
        "source_message": {
          "type": "string",
          "minLength": 50,
          "description": "Your response body (Markdown, ≥50 chars)."
        },
        "budget_currency": {
          "enum": [
            "EUR",
            "USD",
            "UAH"
          ],
          "type": "string",
          "description": "Currency of budget_cents (optional)."
        },
        "source_language": {
          "enum": [
            "en",
            "ru",
            "uk",
            "el",
            "de",
            "fr",
            "es",
            "it",
            "nl"
          ],
          "type": "string",
          "default": "en",
          "description": "Hint only — the actual language is auto-detected from your text. Write in the user's own language."
        }
      }
    }
    arguments 57 lines
  • post_proposal_update auth-required never probed

    Post an updated offer (Markdown body + optional new price) or a comment in YOUR OWN response thread on a project. Append-only history; no access pass needed. IMPORTANT: by default the entry is PUBLIC — visible to competing builders. Set secret:true to post it privately (only the client and you see it): if the response isn't already secret this returns a checkout_url (pay by card in EUR) instead of posting, and after payment the version is published private. Before relying on that checkout_url, offer a promo code: if the user has one, pass promo_code — a valid code makes the secret version free.

    mcp-tool

    {
      "type": "object",
      "required": [
        "project_id",
        "body_markdown"
      ],
      "properties": {
        "kind": {
          "enum": [
            "proposal",
            "comment"
          ],
          "type": "string",
          "default": "proposal",
          "description": "'proposal' = updated offer (default), 'comment' = plain Q&A."
        },
        "secret": {
          "type": "boolean",
          "default": false,
          "description": "Post privately so competitors can't see it (returns a checkout_url — pay by card in EUR — unless the response is already secret; unavailable while payments are switched off)."
        },
        "project_id": {
          "type": "string",
          "description": "Project UUID from list_projects."
        },
        "promo_code": {
          "type": "string",
          "description": "Optional promo code for a secret version — a valid code publishes it privately for free (no checkout_url). Ask the user before paying."
        },
        "budget_cents": {
          "type": "integer",
          "minimum": 1,
          "description": "New proposed budget in minor units (optional; proposals only)."
        },
        "body_markdown": {
          "type": "string",
          "maxLength": 8000,
          "minLength": 1,
          "description": "The updated offer in Markdown (put links/images inline). 1–8000 chars."
        },
        "budget_currency": {
          "enum": [
            "EUR",
            "USD",
            "UAH"
          ],
          "type": "string",
          "description": "Currency of budget_cents (optional)."
        }
      }
    }
    arguments 51 lines
  • get_response_thread auth-required never probed

    Read the timeline of YOUR OWN response on a project — comments and proposal versions (including your private secret-offer entries).

    mcp-tool

    {
      "type": "object",
      "required": [
        "project_id"
      ],
      "properties": {
        "project_id": {
          "type": "string",
          "description": "Project UUID from list_projects."
        }
      }
    }
    arguments 12 lines
  • get_conversation auth-required never probed

    Read one conversation's message timeline (participant only). Messages are shown in YOUR language (auto-translated). can_reply tells you whether you may send yet — a builder must have responded to the project first.

    mcp-tool

    {
      "type": "object",
      "required": [
        "conversation_id"
      ],
      "properties": {
        "conversation_id": {
          "type": "string",
          "description": "From list_conversations."
        }
      }
    }
    arguments 12 lines
  • send_message auth-required never probed

    Send a message in a conversation you are part of. A builder may write only after responding to the project. The message is auto-translated to the recipient's language (spends your AI tokens if the languages differ). ALWAYS confirm the text with the user before sending — this posts on their behalf.

    mcp-tool

    {
      "type": "object",
      "required": [
        "conversation_id",
        "body"
      ],
      "properties": {
        "body": {
          "type": "string",
          "maxLength": 4000,
          "minLength": 1,
          "description": "Message text (1–4000 chars)."
        },
        "conversation_id": {
          "type": "string",
          "description": "From list_conversations."
        }
      }
    }
    arguments 19 lines
  • mark_conversation_read auth-required never probed

    Mark the other party's messages in a conversation as read (clears its unread_count).

    mcp-tool

    {
      "type": "object",
      "required": [
        "conversation_id"
      ],
      "properties": {
        "conversation_id": {
          "type": "string",
          "description": "From list_conversations."
        }
      }
    }
    arguments 12 lines
  • offer_project auth-required never probed

    CLIENT: offer your project to a builder (pick one from list_responses). No contacts are shared yet — the builder must accept_project to close the deal. You can re-offer to another builder until then (the previous offer is cancelled + they are notified).

    mcp-tool

    {
      "type": "object",
      "required": [
        "project_id",
        "builder"
      ],
      "properties": {
        "builder": {
          "type": "string",
          "description": "The builder's @handle to offer (from list_responses)."
        },
        "project_id": {
          "type": "string",
          "description": "Your project UUID."
        }
      }
    }
    arguments 17 lines
  • accept_project auth-required never probed

    BUILDER: accept the client's offer on a project you were offered. The project settles and contacts are exchanged (check your Messages / get_profile flow).

    mcp-tool

    {
      "type": "object",
      "required": [
        "project_id"
      ],
      "properties": {
        "project_id": {
          "type": "string",
          "description": "Project UUID you were offered."
        }
      }
    }
    arguments 12 lines
  • decline_project auth-required never probed

    BUILDER: decline the client's offer on a project you were offered. The project reopens and the client is notified.

    mcp-tool

    {
      "type": "object",
      "required": [
        "project_id"
      ],
      "properties": {
        "project_id": {
          "type": "string",
          "description": "Project UUID you were offered."
        }
      }
    }
    arguments 12 lines
  • buy_service auth-required never probed

    Buy a GPTaria shop service and return a checkout_url; the human opens it and pays by card in EUR through Paddle (the seller of record; VAT is added at checkout where it applies), and it activates automatically once paid. PAID CHECKOUT IS SWITCHED OFF FOR NOW — this returns payments_unavailable; tell the user plainly that paid extras are coming soon and that publishing and responding (with the free Day Pass) cost nothing today. You are a helpful service: once payments are live, use this only when a paid option genuinely moves the user's task forward (an action is blocked for lack of a pass, or the user asks to unlock something). When it does, name the option that fits and point out the better-value 30-day pass over a single Day Pass (call list_services for the exact € prices). And when a builder is posting a response, remind them that a highly detailed, valuable response can be made secret (create_response with secret:true) so competitors can't copy it. Mention plainly and unobtrusively that this is paid, then let the user decide. The user always confirms the payment. Services: day_pass (respond to projects for 1 day) · pass_30 (respond for 30 days — best value for an active builder) · ai_pack_s|m|l (top up AI tokens — adds tokens to your single balance, which is then deducted per AI operation like generating a case summary; buy these when an action fails with need_ai_budget — check get_balance for the current balance) · secret_mode (CLIENT: make YOUR open project private — needs target = the project's UUID or slug) · case_verification (BUILDER: get YOUR case audited + a Verified badge — needs target = the case UUID or slug; the case must already exist and be unverified). ALWAYS confirm with the user before creating a checkout, and offer a promo code first: a valid promo_code activates the service for FREE (no checkout_url). NOTE: publishing a project is create_project; a secret response is create_response(secret:true) — those separate tools carry their own checkout.

    mcp-tool

    {
      "type": "object",
      "required": [
        "service_id"
      ],
      "properties": {
        "target": {
          "type": "string",
          "description": "Required for secret_mode (your project UUID/slug) and case_verification (your case UUID/slug). Omit for passes / AI packs."
        },
        "promo_code": {
          "type": "string",
          "description": "Optional promo code — a valid code activates the service for free (no checkout_url). Ask the user before paying."
        },
        "service_id": {
          "enum": [
            "day_pass",
            "pass_30",
            "ai_pack_s",
            "ai_pack_m",
            "ai_pack_l",
            "secret_mode",
            "case_verification"
          ],
          "type": "string",
          "description": "What to buy."
        }
      }
    }
    arguments 29 lines
  • list_responses auth-required never probed

    List a project's responses (proposals). Content comes back in the account language; pass `language` to override.

    mcp-tool

    {
      "type": "object",
      "required": [
        "project_id"
      ],
      "properties": {
        "language": {
          "enum": [
            "en",
            "ru",
            "uk",
            "el",
            "de",
            "fr",
            "es",
            "it",
            "nl"
          ],
          "type": "string",
          "description": "Render content in this language (default: the account's preferred language)."
        },
        "project_id": {
          "type": "string",
          "description": "Project UUID from list_projects."
        }
      }
    }
    arguments 27 lines
  • list_project_comments auth-required never probed

    List a client's clarifications on a project.

    mcp-tool

    {
      "type": "object",
      "required": [
        "project_id"
      ],
      "properties": {
        "project_id": {
          "type": "string",
          "description": "Project UUID from list_projects."
        }
      }
    }
    arguments 12 lines
  • create_project_comment auth-required never probed

    Add a public clarification to your own project brief (owner only).

    mcp-tool

    {
      "type": "object",
      "required": [
        "project_id",
        "body"
      ],
      "properties": {
        "body": {
          "type": "string",
          "maxLength": 4000,
          "minLength": 1,
          "description": "The clarification text (1–4000 chars)."
        },
        "project_id": {
          "type": "string",
          "description": "Your project UUID."
        }
      }
    }
    arguments 19 lines
  • list_conversations auth-required 4h ago

    List YOUR message conversations (client↔builder, one per project). Each shows the project, the other party (@handle), unread_count and the last message. This is the inbox the "check your Messages" hints (e.g. after accept_project) point to.

    mcp-tool

    {
      "type": "object",
      "properties": {}
    }
    arguments 4 lines
  • create_case auth-required never probed

    Publish a portfolio case (dev.to-style Markdown). Draft by default.

    mcp-tool

    {
      "type": "object",
      "required": [
        "title",
        "body_markdown"
      ],
      "properties": {
        "tags": {
          "type": "array",
          "items": {
            "type": "string",
            "description": "A tag."
          },
          "description": "Tags for the case (optional)."
        },
        "title": {
          "type": "string",
          "maxLength": 120,
          "minLength": 3,
          "description": "Case title (3–120 chars)."
        },
        "published": {
          "type": "boolean",
          "default": false,
          "description": "Publish immediately (default false = draft)."
        },
        "description": {
          "type": "string",
          "maxLength": 300,
          "description": "Short summary (≤300 chars, optional)."
        },
        "body_markdown": {
          "type": "string",
          "minLength": 1,
          "description": "Case body in Markdown."
        },
        "source_language": {
          "enum": [
            "en",
            "ru",
            "uk",
            "el",
            "de",
            "fr",
            "es",
            "it",
            "nl"
          ],
          "type": "string",
          "default": "en",
          "description": "Hint only — the actual language is auto-detected from the body. Write in the user's own language."
        }
      }
    }
    arguments 54 lines
  • list_services auth-required never probed

    The GPTaria shop price list — every service with its price in EUR (€, VAT added at checkout where it applies), what it does, how to buy it, plus tokens per AI pack and days per access pass. FREE. Use it to name and compare plans when you advise — e.g. a single Day Pass vs the better-value 30-day Pass — so you can quote the real € amount. payments_live:false means checkout is switched off for now: say the paid extras are coming soon.

    mcp-tool

    {
      "type": "object",
      "properties": {}
    }
    arguments 4 lines
  • generate_case_summary auth-required never probed

    Generate a case's AI summary block from its body (owner only).

    mcp-tool

    {
      "type": "object",
      "required": [
        "case_id"
      ],
      "properties": {
        "case_id": {
          "type": "string",
          "description": "Case UUID (from create_case or your cases)."
        },
        "author_theses": {
          "type": "string",
          "maxLength": 1000,
          "description": "Optional: what to emphasize (grounds the model, invents nothing)."
        }
      }
    }
    arguments 17 lines
  • update_profile auth-required never probed

    Edit your own GPTaria profile (partial update — only the fields you pass change). Besides the public profile this also sets two account settings the user would otherwise have to click through the website for: `preferred_language` (the language GPTaria writes to them in — emails, site and Telegram notifications) and `notification_categories` (which project categories they get new-project alerts about — this is how a builder says "tell me when a project appears in X"). Read the current state with get_profile first; notification_categories REPLACES the whole set, so include the categories they already have unless they want them off.

    mcp-tool

    {
      "type": "object",
      "properties": {
        "bio": {
          "type": "string",
          "maxLength": 500,
          "description": "Short bio (≤500 chars), in ONE language — auto-translated."
        },
        "handle": {
          "type": "string",
          "pattern": "^[a-z0-9_]{3,20}$",
          "description": "URL handle: 3–20 chars, lowercase letters, digits, underscore."
        },
        "contact_x": {
          "type": "string",
          "maxLength": 100,
          "description": "Public X (Twitter) handle."
        },
        "country_code": {
          "type": "string",
          "maxLength": 2,
          "minLength": 2,
          "description": "2-letter ISO country code, e.g. \"UA\"."
        },
        "display_name": {
          "type": "string",
          "maxLength": 80,
          "description": "Your public display name (≤80 chars)."
        },
        "github_handle": {
          "type": "string",
          "maxLength": 60,
          "description": "GitHub username (with or without @)."
        },
        "contact_linkedin": {
          "type": "string",
          "maxLength": 200,
          "description": "Public LinkedIn handle or URL."
        },
        "contact_telegram": {
          "type": "string",
          "maxLength": 100,
          "description": "Public Telegram handle."
        },
        "working_languages": {
          "type": "array",
          "items": {
            "type": "string",
            "pattern": "^[a-z]{2}$",
            "description": "2-letter language code, e.g. \"en\"."
          },
          "description": "Languages you work in (≤8)."
        },
        "preferred_language": {
          "enum": [
            "en",
            "ru",
            "uk",
            "el",
            "de",
            "fr",
            "es",
            "it",
            "nl"
          ],
          "type": "string",
          "description": "Language GPTaria writes to the user in: emails, site and Telegram notifications. Not the same as bio_source_language (that describes one text)."
        },
        "bio_source_language": {
          "enum": [
            "en",
            "ru",
            "uk",
            "el",
            "de",
            "fr",
            "es",
            "it",
            "nl"
          ],
          "type": "string",
          "description": "Hint only — the bio's language is auto-detected from its text. Write the bio in the user's own language."
        },
        "notification_categories": {
          "type": "array",
          "items": {
            "enum": [
              "software",
              "devops",
              "video",
              "audio",
              "image",
              "content",
              "marketing",
              "social",
              "other"
            ],
            "type": "string",
            "description": "Project category slug."
          },
          "description": "Get an email + Telegram alert whenever a new project appears in these categories (they also decide where clients find the user in the catalog). REPLACES the whole set; [] turns every alert off."
        }
      }
    }
    arguments 104 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.

_ for your README measured, not declared

measured by brick.blue

[![measured by brick.blue](https://brick.blue/api/v1/agents/eefcfe49f00c4d97/badge.svg)](https://brick.blue/agent/eefcfe49f00c4d97)

The picture says what this hub measured — the access class, how many tools it called and whether they answered — and refreshes hourly. Own the domain? Prove it and the listing carries a verified badge here too: passport.

_ 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.