Server definition

Hash: sha256:b4ee406721cff36648081c4cbddfa3a52075977ff5b27e62d8a10d224fe55db0
What it is: What a remote MCP server returned when asked what it offers: 16 tools

The blob, as servednamed by its sha256

{
  "instructions": "Maginary (spelled M-A-G-I-N-A-R-Y, not 'imaginary') is a Midjourney-style AI image + video generator with a `--flag` prompt DSL and an async HTTP API.\n\n**New users (no API key yet) — pick the shortest path:**\n- **USDC wallet (fastest):** Just call `generate` with your prompt. The 402 response prices the generation; settle it on-chain (x402, USDC on Base) and your account is created automatically.\n- **Wallet auth (no email):** `create_wallet_account` → returns an API key immediately, no email verification.\n- **Email:** `create_account` → user clicks verification email → `manage_api_key(action='create')` → `configure_api_key` → `get_products` → `checkout`. The $10 novice_pack is the recommended starting point. API key creation and checkout require a verified email.\n\n**Prompt rules (IMPORTANT):** Pass the user's words as-is. Do NOT add flags the user did not ask for — no ``--ar``, no ``--flagship``, no model flags, nothing. Flags cost money; adding them without asking is a UX failure. Before the FIRST generation in a conversation, briefly mention the quality option once: standard (default, cheap) or flagship (best, ~4× more expensive) — then generate with whatever they choose, or standard if they don't care. After that, never ask again unless they bring it up. Do NOT ask about aspect ratio — omit ``--ar`` by default; square format gives the most diversity and is the best starting point. Only add ``--ar`` when the user explicitly mentions landscape, portrait, widescreen, etc. Flags go at the END of the prompt. ``--1``/``--2``/``--3``/``--4`` = image count (default 4; only if user specifies). ``--ar 16:9`` = aspect ratio (only if user explicitly asks). ``--flagship`` = best quality (only if user explicitly wants it). Unknown flag: call ``get_parameter(name)`` — never guess.\n\n**Every flag that exists, and its state:** Flags, live (35): --ar, --output-count (--1/--2/--3/--4), --seed, --transparent, --sref, --sw, --png, --jpg, --webp, --svg, --2k, --4k, --upscale, --vary, --varysubtle, --varystrong, --panleft, --panright, --panup, --pandown, --zoomout, --mp4, --video-resolution (--480p/--540p/--720p/--1024p/--1080p/--2160p / --4k (4k, Seedance 2 Pro)/--480p24 / --480p24fps/--540p24 / --540p24fps/--720p24 / --720p24fps/--1024p30 / --1024p30fps/--1080p24 / --1080p24fps), --video-fps (--24fps/--30fps/--50fps/--60fps), --video-duration (--4s / --4sec/--5s / --5sec/--6s / --6sec/--8s / --8sec/--10s / --10sec/--12s / --12sec), --flagship, --sora, --soralite, --nanobananapro, --nb2, --gpt2, --gpt2high, --seedance2, --seedance2pro, --demo. Partial (4, only some models honour them): --no, --zoomout2x, --zoomoutexpand, --zoomoutexpand2x. Reserved (2, the parser rejects them): --cref, --cw. Any other --flag is rejected with `Unrecognized parameter`. Details: `get_parameter(name)`.\n\n**Existing users:** Use `search_parameters` / `get_parameter` to discover which flags exist before building a prompt. Use `generate` to kick off a generation, then `wait_for_generation` (or a webhook callback) to fetch the resulting image / video URLs. Flags whose status is `dead`, `mostly-dead`, or `unimplemented` should be avoided.\n\n**Image editing (img2img):** Place image URLs directly in the prompt (e.g. `https://cdn.maginary.ai/…/photo.webp reimagine as watercolor`). If the user already has a URL, use it directly. On local (stdio) connections, `upload_image(file_path=...)` reads a file from disk and returns a CDN URL. `--sref <url>` is style-only, not img2img.\n\n**Follow-up actions:** A completed generation's `processing_result.available_actions` lists what's available (upscale, vary, pan, zoom, img2vid, reroll). Use `execute_action` with the `generation_uuid`, `action_type`, and `parent_image_index`.\n\n**Credits:** When the user asks to generate, just call `generate` — don't check balance first. A zero balance does not mean the user cannot generate (x402 on-chain payment settles inline). The server returns a clear `payment_required` response when credits are needed.",
  "tools": [
    {
      "description": "Check account verification status, credit balance, and API key count.\n\n    Use this after ``create_account`` to poll whether the user has clicked the\n    verification link. Pass ``email`` + ``password`` (from ``create_account``)\n    for Basic auth, or omit both to use the configured API key.\n\n    Args:\n        email: Account email (for Basic auth).\n        password: Account password (for Basic auth).\n\n    Returns:\n        Dict with ``verified`` (bool), ``email``, ``api_key_count``,\n        ``credits_remaining``, ``uploads_remaining``.\n    ",
      "inputSchema": {
        "properties": {
          "email": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Account email (for Basic auth). Omit to use API key.",
            "title": "Email"
          },
          "password": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Account password (for Basic auth).",
            "title": "Password"
          }
        },
        "title": "check_account_statusArguments",
        "type": "object"
      },
      "name": "check_account_status",
      "outputSchema": {
        "additionalProperties": true,
        "title": "check_account_statusDictOutput",
        "type": "object"
      }
    },
    {
      "description": "Create a Stripe checkout session for purchasing a product.\n\n    Returns a ``checkout_url`` — the user must open it in a browser to\n    complete payment. After payment, credits are provisioned automatically\n    via webhook. **Present the URL exactly as returned, including the\n    ``#fragment`` — do not truncate, reformat, or strip any part of it.**\n\n    If the agent has a USDC wallet, skip this entirely — just call\n    ``generate`` and the x402 protocol handles payment on-chain.\n\n    Args:\n        product_id: Product ID from ``get_products``.\n        email: Account email (for Basic auth during onboarding).\n        password: Account password (for Basic auth during onboarding).\n\n    Returns:\n        Dict with ``checkout_url``. On failure, an ``isError`` result — e.g.\n        ``error: \"email_not_verified\"`` until the user clicks the\n        verification link, or ``\"auth\"`` / ``\"failed\"``.\n    ",
      "inputSchema": {
        "properties": {
          "email": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Account email (for Basic auth during onboarding).",
            "title": "Email"
          },
          "password": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Account password (for Basic auth).",
            "title": "Password"
          },
          "product_id": {
            "description": "Product ID from get_products.",
            "title": "Product Id",
            "type": "integer"
          }
        },
        "required": [
          "product_id"
        ],
        "title": "checkoutArguments",
        "type": "object"
      },
      "name": "checkout",
      "outputSchema": {
        "additionalProperties": true,
        "title": "checkoutDictOutput",
        "type": "object"
      }
    },
    {
      "description": "Activate an API key. Local (stdio) servers persist it; hosted does not.\n\n    Call this after ``manage_api_key(action='create')`` returns a ``raw_key``.\n    On a local server the key is saved to ``~/.config/maginary/api_key``\n    (chmod 600) and survives restarts. On the hosted server\n    (mcp.maginary.ai) nothing can be stored — auth is per-request: the\n    response will say ``persisted: false`` and the key must be sent as an\n    ``Authorization: Bearer <key>`` header on every request (set it in the\n    MCP client's connection config).\n\n    Args:\n        api_key: The full API key string returned by ``manage_api_key``.\n\n    Returns:\n        Confirmation dict.\n    ",
      "inputSchema": {
        "properties": {
          "api_key": {
            "description": "Full API key string from manage_api_key.",
            "title": "Api Key",
            "type": "string"
          }
        },
        "required": [
          "api_key"
        ],
        "title": "configure_api_keyArguments",
        "type": "object"
      },
      "name": "configure_api_key",
      "outputSchema": {
        "additionalProperties": true,
        "title": "configure_api_keyDictOutput",
        "type": "object"
      }
    },
    {
      "description": "Create a new Maginary account for the given email address.\n\n    Returns the auto-generated password — display it to the user ONCE so they\n    can save it. A verification email is sent; the user must click the link\n    before the account can generate images.\n\n    After verification, use ``manage_api_key(action='create')`` with\n    ``email`` + ``password`` to get an API key, then ``configure_api_key``\n    to activate it.\n\n    Args:\n        email: The user's email address.\n\n    Returns:\n        Dict with ``email``, ``password``, and ``message``. On failure, an\n        ``isError`` result — e.g. ``error: \"already_exists\"`` (email taken:\n        ask the user for their password or a different email),\n        ``\"rate_limited\"``, or ``\"failed\"``.\n    ",
      "inputSchema": {
        "properties": {
          "email": {
            "description": "Email address for the new account.",
            "title": "Email",
            "type": "string"
          }
        },
        "required": [
          "email"
        ],
        "title": "create_accountArguments",
        "type": "object"
      },
      "name": "create_account",
      "outputSchema": {
        "additionalProperties": true,
        "title": "create_accountDictOutput",
        "type": "object"
      }
    },
    {
      "description": "Create (or access) a Maginary account using a wallet signature.\n\n    Sign the message ``Maginary: authenticate <address> at <timestamp>.\n    This does not move funds.`` with EIP-191 ``personal_sign`` and pass all\n    three values. On success, an API key is returned immediately — no email\n    verification needed.\n\n    Use this when you have a wallet but no email. The returned ``api_key``\n    should be passed as ``Authorization: Bearer <key>`` in the MCP client\n    config, or via ``configure_api_key`` (stdio) / ``_meta[\"maginary/api_key\"]``\n    (hosted, per-call).\n\n    If the wallet already has an account, returns the existing account with\n    a fresh API key.\n\n    Args:\n        address: EVM wallet address (0x..., 42 chars).\n        signature: Hex-encoded EIP-191 personal_sign of the auth message.\n        timestamp: Unix epoch seconds used in the signed message (must be\n            within the last 5 minutes).\n\n    Returns:\n        Dict with ``address``, ``api_key`` (full key — show once),\n        ``key_prefix``, ``created`` (bool), ``message``.\n        On failure: ``isError`` with ``error`` = ``\"validation\"``,\n        ``\"signature_failed\"``, or ``\"rate_limited\"``.\n    ",
      "inputSchema": {
        "properties": {
          "address": {
            "description": "EVM wallet address (0x..., 42 chars).",
            "title": "Address",
            "type": "string"
          },
          "signature": {
            "description": "Hex EIP-191 personal_sign of the auth message.",
            "title": "Signature",
            "type": "string"
          },
          "timestamp": {
            "description": "Unix epoch seconds used in the signed message.",
            "title": "Timestamp",
            "type": "integer"
          }
        },
        "required": [
          "address",
          "signature",
          "timestamp"
        ],
        "title": "create_wallet_accountArguments",
        "type": "object"
      },
      "name": "create_wallet_account",
      "outputSchema": {
        "additionalProperties": true,
        "title": "create_wallet_accountDictOutput",
        "type": "object"
      }
    },
    {
      "description": "Run a follow-up action on a completed generation's image.\n\n    After ``generate`` → ``wait_for_generation``, the response's\n    ``processing_result.available_actions`` lists what's possible per slot.\n    Call this tool with one of those action types.\n\n    Args:\n        generation_uuid: UUID of the parent generation (from ``generate``).\n        action_type: One of the values from ``available_actions`` — e.g.\n            ``\"upscale_2x\"``, ``\"upscale_1_5x\"``, ``\"vary_strong\"``,\n            ``\"vary_subtle\"``, ``\"pan_left\"``, ``\"pan_right\"``,\n            ``\"pan_up\"``, ``\"pan_down\"``, ``\"zoom_out_2x\"``,\n            ``\"zoom_out_1_5x\"``, ``\"img2vid_basic\"``, ``\"reroll\"``.\n        parent_image_index: The slot index of the image to act on (0, 1,\n            2, or 3 for a 4-image grid). Required for per-slot actions;\n            omit for ``\"reroll\"`` (global action).\n        prompt: Optional replacement prompt. For ``vary_*`` you can steer\n            the variation with a new prompt; for ``img2vid_basic`` you can\n            describe the desired motion.\n        callback_url: Optional webhook URL (same as ``generate``).\n\n    Returns:\n        The newly created child generation record (same shape as\n        ``generate``'s return — poll it with ``wait_for_generation``).\n\n        On failure, same ``isError`` contract as ``generate``:\n        ``\"auth\"``, ``\"payment_required\"`` (with x402 challenge),\n        or ``\"failed\"``.\n    ",
      "inputSchema": {
        "properties": {
          "action_type": {
            "description": "Action from available_actions, e.g. upscale_2x, vary_strong, img2vid_basic, reroll.",
            "title": "Action Type",
            "type": "string"
          },
          "callback_url": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "HTTPS webhook URL for done/failed notifications.",
            "title": "Callback Url"
          },
          "generation_uuid": {
            "description": "UUID of the parent generation.",
            "title": "Generation Uuid",
            "type": "string"
          },
          "parent_image_index": {
            "anyOf": [
              {
                "type": "integer"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Slot index (0-3) of the image to act on. Omit for global actions like reroll.",
            "title": "Parent Image Index"
          },
          "prompt": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Optional replacement prompt for vary/img2vid actions.",
            "title": "Prompt"
          }
        },
        "required": [
          "generation_uuid",
          "action_type"
        ],
        "title": "execute_actionArguments",
        "type": "object"
      },
      "name": "execute_action",
      "outputSchema": {
        "additionalProperties": true,
        "title": "execute_actionDictOutput",
        "type": "object"
      }
    },
    {
      "description": "Kick off a generation via POST /api/gens/.\n\n    Args:\n        prompt: The user's words, passed through as-is. Do NOT add flags the\n            user did not ask for — no ``--ar``, no ``--flagship``, no model\n            flags. Every extra flag costs credits; adding them unrequested is\n            wrong. Standard quality is the default and is cheap; ``--flagship``\n            is ~4× more expensive and must only be used when the user\n            explicitly asks for best quality.\n\n            If the user asks about quality or aspect ratio: ask them first\n            (standard vs flagship, landscape vs portrait) before generating.\n\n            Flags go at the END, only when the user asked:\n            ``--1``/``--2``/``--3``/``--4`` = image count (default 4),\n            ``--ar 16:9`` = aspect ratio, ``--flagship`` = best quality.\n            Unknown flag: call ``get_parameter(name)`` first — never guess.\n\n            Examples — user says \"a fox\": prompt is ``\"a fox\"``.\n            User says \"a fox, landscape, best quality\":\n            prompt is ``\"a fox --ar 16:9 --flagship\"``.\n\n            **Image-to-image (img2img):** Place one or more public image URLs\n            in the prompt, followed by editing instructions:\n            ``\"https://cdn.example.com/photo.webp reimagine as oil painting --ar 16:9\"``\n            The engine extracts URLs automatically and switches to img2img mode.\n            Multiple URLs trigger multi-input mode (compositing/combining).\n            Use ``upload_image`` first if images aren't already hosted.\n\n            **Image-to-video:** Place an image URL in the prompt AND add\n            ``--mp4`` plus video flags (``--5sec``, ``--1080p``). Or use\n            ``execute_action`` with ``action_type=\"img2vid_basic\"`` on a\n            completed generation's image.\n\n            **Style reference (--sref) is NOT img2img:** ``--sref <url>``\n            copies the visual *style* of a reference image (colors, mood,\n            composition) without using the image content as input. A bare URL\n            in the prompt edits the actual image; ``--sref`` transfers style.\n\n        callback_url: Optional HTTPS URL that will receive a webhook when the\n            generation reaches done / failed. See\n            https://maginary.ai/blog/webhooks-guide for signature verification.\n\n    Returns:\n        On success, the created generation record. Key fields: ``uuid`` (use\n        to poll), ``action_type``, ``processing_state``,\n        ``expected_output_count``.\n\n        On failure, an ``isError`` result instead (nothing is raised), with a\n        JSON body whose ``error`` field is one of:\n\n        - ``\"auth\"`` — no/invalid API key. Surface the message directly to\n          the human.\n        - ``\"payment_required\"`` — out of credits. The body carries\n          ``billing_url`` and top-level x402 fields (``accepts``,\n          ``resource``): either send the human to ``billing_url`` to top up,\n          or pay programmatically via x402 (settle ``accepts[0]`` with USDC\n          on Base and retry).\n        - ``\"demo_not_found\"`` — ``--demo`` prompt has no matching seeded\n          generation. ``available_demos`` lists valid prompts.\n        - ``\"failed\"`` — anything else (invalid prompt, rate limit, backend\n          or network error); see ``message``.\n\n        x402 over MCP: a ``payment_required`` result also carries the x402\n        fields at the top level (``accepts``, ``resource``); an x402-capable\n        client signs ``accepts[0]`` and calls this tool again with the payment\n        in ``_meta[\"x402/payment\"]``. The settled call returns the generation\n        with ``x402_receipt`` (and ``_meta[\"x402/payment-response\"]``); a\n        wallet's first settlement creates its account. Subsequent requests\n        use wallet-signed auth headers (X-Wallet-Address/Signature/Timestamp)\n        or pass an API key as ``_meta[\"maginary/api_key\"]``.\n\nEvery flag that exists, and its state: Flags, live (35): --ar, --output-count (--1/--2/--3/--4), --seed, --transparent, --sref, --sw, --png, --jpg, --webp, --svg, --2k, --4k, --upscale, --vary, --varysubtle, --varystrong, --panleft, --panright, --panup, --pandown, --zoomout, --mp4, --video-resolution (--480p/--540p/--720p/--1024p/--1080p/--2160p / --4k (4k, Seedance 2 Pro)/--480p24 / --480p24fps/--540p24 / --540p24fps/--720p24 / --720p24fps/--1024p30 / --1024p30fps/--1080p24 / --1080p24fps), --video-fps (--24fps/--30fps/--50fps/--60fps), --video-duration (--4s / --4sec/--5s / --5sec/--6s / --6sec/--8s / --8sec/--10s / --10sec/--12s / --12sec), --flagship, --sora, --soralite, --nanobananapro, --nb2, --gpt2, --gpt2high, --seedance2, --seedance2pro, --demo. Partial (4, only some models honour them): --no, --zoomout2x, --zoomoutexpand, --zoomoutexpand2x. Reserved (2, the parser rejects them): --cref, --cw. Any other --flag is rejected with `Unrecognized parameter`. Details: `get_parameter(name)`.",
      "inputSchema": {
        "properties": {
          "callback_url": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "HTTPS webhook URL for done/failed notifications.",
            "title": "Callback Url"
          },
          "prompt": {
            "description": "The user's words as-is, flags at the end. Do NOT add flags the user did not ask for.",
            "title": "Prompt",
            "type": "string"
          }
        },
        "required": [
          "prompt"
        ],
        "title": "generateArguments",
        "type": "object"
      },
      "name": "generate",
      "outputSchema": {
        "additionalProperties": true,
        "title": "generateDictOutput",
        "type": "object"
      }
    },
    {
      "description": "Check remaining credits and uploads for the authenticated account.\n\n    Args:\n        email: Account email (for Basic auth).\n        password: Account password (for Basic auth).\n\n    Returns:\n        Dict with ``credits_remaining`` and ``uploads_remaining``.\n    ",
      "inputSchema": {
        "properties": {
          "email": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Account email (for Basic auth).",
            "title": "Email"
          },
          "password": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Account password (for Basic auth).",
            "title": "Password"
          }
        },
        "title": "get_balanceArguments",
        "type": "object"
      },
      "name": "get_balance",
      "outputSchema": {
        "additionalProperties": true,
        "title": "get_balanceDictOutput",
        "type": "object"
      }
    },
    {
      "description": "Fetch a generation by UUID (GET /api/gens/{uuid}/).\n\n    Args:\n        uuid: The UUID returned by ``generate``.\n\n    Returns:\n        The full generation record. If terminal, ``image_urls[]`` holds the\n        finished outputs and ``processing_result.slots[]`` the per-slot detail.\n        NOTE: a generation that failed server-side is a SUCCESSFUL tool call\n        returning ``processing_state: \"failed\"`` — always check the state,\n        never infer success from the absence of a tool error.\n\n        **Follow-up actions:** A completed generation's\n        ``processing_result.available_actions`` maps slot indices to valid\n        action types. E.g. ``{\"0\": [\"upscale_2x\", \"vary_strong\", ...],\n        \"global\": [\"reroll\"]}``. Use ``execute_action`` with the ``uuid``,\n        a chosen ``action_type``, and the ``parent_image_index`` (the slot\n        key as an int) to run an action.\n\n        Hosted: a key obtained mid-session may be passed as\n        ``_meta[\"maginary/api_key\"]``.\n    ",
      "inputSchema": {
        "properties": {
          "uuid": {
            "description": "Generation UUID from generate or execute_action.",
            "title": "Uuid",
            "type": "string"
          }
        },
        "required": [
          "uuid"
        ],
        "title": "get_generationArguments",
        "type": "object"
      },
      "name": "get_generation",
      "outputSchema": {
        "additionalProperties": true,
        "title": "get_generationDictOutput",
        "type": "object"
      }
    },
    {
      "description": "Return the full record for a single parameter (canonical name or alias).\n\n    Args:\n        name: Parameter name with or without leading ``--`` (e.g. ``ar``,\n            ``--ar``, ``aspect``). Case-insensitive.\n\n    Returns:\n        The parameter dict. Not-found is an ``isError`` result — surface it\n        rather than fabricating a param.\n    ",
      "inputSchema": {
        "properties": {
          "name": {
            "description": "Parameter name with or without --, e.g. ar, --ar, aspect.",
            "title": "Name",
            "type": "string"
          }
        },
        "required": [
          "name"
        ],
        "title": "get_parameterArguments",
        "type": "object"
      },
      "name": "get_parameter",
      "outputSchema": {
        "additionalProperties": true,
        "title": "get_parameterDictOutput",
        "type": "object"
      }
    },
    {
      "description": "List available Maginary products/plans with pricing.\n\n    No authentication required. Use this to present purchase options to the\n    user. The ``novice_pack`` ($10, 150 credits) is the recommended starting\n    point.\n\n    Returns:\n        Dict with ``count`` and ``products`` — each product carries ``id``,\n        ``short_name``, ``title``, ``description``, ``price_cents``,\n        ``credits``, ``uploads``, ``is_subscription``. (The backend sends a\n        bare array; it is wrapped here because FastMCP validates tool output\n        against the dict annotation and rejects a top-level list.)\n    ",
      "inputSchema": {
        "properties": {},
        "title": "get_productsArguments",
        "type": "object"
      },
      "name": "get_products",
      "outputSchema": {
        "additionalProperties": true,
        "title": "get_productsDictOutput",
        "type": "object"
      }
    },
    {
      "description": "List Maginary prompt-DSL parameters.\n\n    Args:\n        category: Restrict to one category (e.g. ``composition``, ``video``,\n            ``model``, ``outpaint``). Call with no filters once — the response's\n            ``categories`` / ``statuses`` maps are the full taxonomy.\n        status: Restrict to one status (``live``, ``mostly-dead``,\n            ``unimplemented``).\n        include_reserved: When False (default) drop ``unimplemented``\n            (recognized-but-blocked) parameters from the result.\n\n    Returns:\n        A dict with ``count``, ``source`` (``live`` vs. ``bundled-snapshot``),\n        ``categories`` / ``statuses`` (the filter taxonomy), and ``parameters``\n        (the array of matching entries).\n    ",
      "inputSchema": {
        "properties": {
          "category": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Filter by category, e.g. composition, video, model, outpaint.",
            "title": "Category"
          },
          "include_reserved": {
            "default": false,
            "description": "Include unimplemented (blocked) parameters.",
            "title": "Include Reserved",
            "type": "boolean"
          },
          "status": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Filter by status: live, mostly-dead, or unimplemented.",
            "title": "Status"
          }
        },
        "title": "list_parametersArguments",
        "type": "object"
      },
      "name": "list_parameters",
      "outputSchema": {
        "additionalProperties": true,
        "title": "list_parametersDictOutput",
        "type": "object"
      }
    },
    {
      "description": "Create, list, or revoke Maginary API keys (up to 10 per account).\n\n    Auth: pass ``email`` + ``password`` for Basic auth (onboarding), or omit\n    both to use the configured API key (normal operation).\n\n    Args:\n        action: One of ``create``, ``list``, ``revoke``.\n        name: Key name (required for ``create``).\n        key_prefix: 8-char prefix of the key to revoke (required for ``revoke``).\n        email: Account email (for Basic auth).\n        password: Account password (for Basic auth).\n\n    Returns:\n        For ``create``: dict with ``raw_key`` (the full key — show once, then\n        use ``configure_api_key`` to activate it), ``key_prefix``, ``name``.\n        For ``list``: dict with ``keys`` array.\n        For ``revoke``: success/error message.\n    ",
      "inputSchema": {
        "properties": {
          "action": {
            "description": "One of: create, list, revoke.",
            "title": "Action",
            "type": "string"
          },
          "email": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Account email (for Basic auth).",
            "title": "Email"
          },
          "key_prefix": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "8-char prefix of key to revoke (required for revoke).",
            "title": "Key Prefix"
          },
          "name": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Key name (required for create).",
            "title": "Name"
          },
          "password": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Account password (for Basic auth).",
            "title": "Password"
          }
        },
        "required": [
          "action"
        ],
        "title": "manage_api_keyArguments",
        "type": "object"
      },
      "name": "manage_api_key",
      "outputSchema": {
        "additionalProperties": true,
        "title": "manage_api_keyDictOutput",
        "type": "object"
      }
    },
    {
      "description": "Text-search over parameter names, aliases, descriptions, values, examples.\n\n    Args:\n        query: Substring match, case-insensitive.\n        category: Optional single-category restriction.\n        include_reserved: Whether to include ``unimplemented`` parameters.\n\n    Returns:\n        Dict with ``count``, ``source`` (``live`` vs. ``bundled-snapshot``),\n        and ``parameters`` (ordered as they appear in the catalog).\n    ",
      "inputSchema": {
        "properties": {
          "category": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Filter by category, e.g. composition, video, model.",
            "title": "Category"
          },
          "include_reserved": {
            "default": false,
            "description": "Include unimplemented (blocked) parameters.",
            "title": "Include Reserved",
            "type": "boolean"
          },
          "query": {
            "description": "Search term (case-insensitive substring match).",
            "title": "Query",
            "type": "string"
          }
        },
        "required": [
          "query"
        ],
        "title": "search_parametersArguments",
        "type": "object"
      },
      "name": "search_parameters",
      "outputSchema": {
        "additionalProperties": true,
        "title": "search_parametersDictOutput",
        "type": "object"
      }
    },
    {
      "description": "Upload a local image and get a CDN URL for img2img or ``--sref``.\n\n    Only available on local (stdio) connections.  On hosted/remote\n    connections, place an existing image URL directly in the prompt.\n\n    Place the returned ``url`` in a ``generate`` prompt:\n    ``generate(\"https://cdn.maginary.ai/…/photo.webp reimagine as oil painting\")``\n\n    Args:\n        file_path: Path to an image file on disk (JPEG, PNG, WebP, HEIC).\n        filename: Original filename.  Inferred from ``file_path`` if omitted.\n\n    Returns:\n        Dict with ``url`` (the public CDN URL), ``exists`` (deduplicated),\n        ``credits_deducted``, and ``message``.\n    ",
      "inputSchema": {
        "properties": {
          "file_path": {
            "description": "Path to a local image (JPEG, PNG, WebP, HEIC).",
            "title": "File Path",
            "type": "string"
          },
          "filename": {
            "anyOf": [
              {
                "type": "string"
              },
              {
                "type": "null"
              }
            ],
            "default": null,
            "description": "Override filename. Inferred from file_path if omitted.",
            "title": "Filename"
          }
        },
        "required": [
          "file_path"
        ],
        "title": "upload_imageArguments",
        "type": "object"
      },
      "name": "upload_image",
      "outputSchema": {
        "additionalProperties": true,
        "title": "upload_imageDictOutput",
        "type": "object"
      }
    },
    {
      "description": "Poll ``get_generation`` on a backoff until it reaches done / failed.\n\n    Args:\n        uuid: The UUID returned by ``generate``.\n        timeout_s: Return after this many seconds even if still running.\n            Default 45 stays under the 60 s per-call limit most MCP clients\n            enforce; a ``timeout`` result just means \"call again\". Only raise\n            it (e.g. for video) on clients you know allow long tool calls.\n\n    Returns:\n        The terminal generation record — which includes generations that\n        failed server-side: those are SUCCESSFUL tool calls returning\n        ``processing_state: \"failed\"`` with empty ``image_urls``, so always\n        check the state. On tool failure, an ``isError`` result whose\n        ``error`` field is ``\"timeout\"`` (``message`` names the last\n        observed state — the generation keeps running server-side and can be\n        re-fetched with ``get_generation`` later), ``\"auth\"``, or\n        ``\"failed\"``.\n\n        **Follow-up actions:** A ``done`` generation's\n        ``processing_result.available_actions`` maps slot indices to valid\n        action types — e.g. ``{\"0\": [\"upscale_2x\", \"vary_strong\",\n        \"pan_left\", \"zoom_out_2x\", \"img2vid_basic\", ...], \"global\":\n        [\"reroll\"]}``. Use ``execute_action`` with the ``uuid``, a chosen\n        ``action_type``, and the ``parent_image_index`` (the slot key as an\n        int) to run an action on a specific output image.\n    ",
      "inputSchema": {
        "properties": {
          "timeout_s": {
            "default": 45,
            "description": "Max seconds to wait before returning a timeout result.",
            "title": "Timeout S",
            "type": "number"
          },
          "uuid": {
            "description": "Generation UUID to poll.",
            "title": "Uuid",
            "type": "string"
          }
        },
        "required": [
          "uuid"
        ],
        "title": "wait_for_generationArguments",
        "type": "object"
      },
      "name": "wait_for_generation",
      "outputSchema": {
        "additionalProperties": true,
        "title": "wait_for_generationDictOutput",
        "type": "object"
      }
    }
  ]
}

Verify it yourself

curl -s https://api.teppi.xyz/v1/evidence/sha256:b4ee406721cff36648081c4cbddfa3a52075977ff5b27e62d8a10d224fe55db0 | sha256sum