Yes — it answered the hub's last check (checked 1h ago). It answered 100% of checks over the last 30 days.

Is curated free to use?

Not anonymously — it asks for a key or a login before it will serve. Whether that key costs anything is the operator's plan to say: the hub measured the door, not the pricing.

What tools does curated have?

48 tools: unarchive_list, list_my_lists, update_list, removed_items, create_list, archive_list, delete_list, add_item, ….

Is curated 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.

_ registry / mcp streamable-http · checked 1h ago

curated

https://mcp.curated.network

Registry code: 85470c9c90525ca1

api record

Curated keeps people's recommendations in lists. Full guide, the primary public HTML page: https://curated.network/agent-setup/instructions. Reading it is optional; other formats and the skill are not prerequisites. Continue the person's original task in this conversation. Core workflow: call account_context and confirm the account; create_list for a new list and reuse its exact returned slug; add_item or add_items with the person's own reason in note, never rewritten; read back with get_list_items before reporting success. New lists stay private. Publish, save (save_list, save_item), follow…

endpoint: https://mcp.curated.network
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 curated live?: Yes — it answered the hub's last check (checked 1h ago). It answered 100% of checks over the last 30 days.
Is curated free to use?: Not anonymously — it asks for a key or a login before it will serve. Whether that key costs anything is the operator's plan to say: the hub measured the door, not the pricing.
What tools does curated have?: 48 tools: unarchive_list, list_my_lists, update_list, removed_items, create_list, archive_list, delete_list, add_item, ….
Is curated 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 100%· all time 100%

latency

381ms

last good check

priced tools

of 48 tools

_ answered our checks, 90 days 1 checks · signed record

2026-10-17 12:48 unknown → live

_ 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

through this hub

calls served

successful

paid through this hub

0 USDC

what callers paid

_ what it can do 48 tools

48 auth-required 48 of 48 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.

unarchive_list auth-required never probed

Bring an archived list back. It returns as a private list: going live again is a separate publish the person chooses.

mcp-tool

{
  "type": "object",
  "required": [
    "list"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Which archived list: slug, exact title, or id."
    }
  },
  "additionalProperties": false
}

arguments 13 lines

list_my_lists auth-required 1h ago
Return the person's own lists (slug, title, status, Inbox flag) and the lists they have access to. Call before add_item when unsure which list to target.

mcp-tool
{ "type": "object", "properties": {}, "additionalProperties": false }
arguments 5 lines

update_list auth-required never probed

Edit a list's title, subtitle, category, or the curator's own About answers. Saves the working version without publishing. Readers keep seeing the last approved version until the owner approves a list update.

mcp-tool

{
  "type": "object",
  "required": [
    "list"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Which list: slug, exact title, or id. Must match exactly one list or the call fails."
    },
    "about": {
      "type": "object",
      "properties": {
        "how": {
          "type": "string",
          "maxLength": 1200,
          "description": "The rule for what makes the cut."
        },
        "who": {
          "type": "string",
          "maxLength": 1200,
          "description": "Who the list is and is not for."
        },
        "why": {
          "type": "string",
          "maxLength": 1200,
          "description": "What started this list."
        }
      },
      "description": "The curator's own words about this list. Ask for their answer; do not write personal experience for them.",
      "additionalProperties": false
    },
    "title": {
      "type": "string",
      "description": "New title."
    },
    "category": {
      "type": "string",
      "description": "The category the person chose: Food & Drink, Travel, Books, Music, Movies & TV, Gear & Tools, Software, Fashion, Home & Garden, Ideas & Essays, Art, or Spirituality. Their choice stays; an empty string clears it."
    },
    "subtitle": {
      "type": "string",
      "description": "New subtitle. Pass an empty string to clear."
    }
  },
  "additionalProperties": false
}

arguments 47 lines

removed_items auth-required never probed

List picks removed from one of the person's lists that have not been put back, newest first, with their notes and ids for restore_item. Follow nextCursor as cursor to reach older picks.

mcp-tool

{
  "type": "object",
  "required": [
    "list"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Which list: exact slug, handle/slug, or id from a previous result."
    },
    "cursor": {
      "type": "string",
      "description": "nextCursor from the previous removed_items page; omit for the newest picks."
    }
  },
  "additionalProperties": false
}

arguments 17 lines

create_list auth-required never probed

Create a new private list and return the exact slug to reuse in add_item. Repeating the same request dedupes instead of creating another recent list.

mcp-tool

{
  "type": "object",
  "required": [
    "title"
  ],
  "properties": {
    "title": {
      "type": "string",
      "description": "List title, 120 characters or fewer."
    },
    "category": {
      "type": "string",
      "description": "Usually omit: Curated picks the category from the title and picks. Set it only when the person names one: Food & Drink, Travel, Books, Music, Movies & TV, Gear & Tools, Software, Fashion, Home & Garden, Ideas & Essays, Art, or Spirituality."
    },
    "subtitle": {
      "type": "string",
      "description": "One-line description of what the list is and who it's for."
    },
    "idempotency_key": {
      "type": "string",
      "maxLength": 200,
      "description": "Optional retry key. Reusing it with identical input replays the first result."
    }
  },
  "additionalProperties": false
}

arguments 26 lines

archive_list auth-required never probed
Archive one of the person's lists: it leaves their profile, search and readers' view, and its page stops working. Nothing is lost; unarchive_list brings it back. Confirm with the person first.

mcp-tool
{ "type": "object", "required": [ "list" ], "properties": { "list": { "type": "string", "description": "Which list: slug, exact title, or id. Must match exactly one list or the call fails." } }, "additionalProperties": false }
arguments 13 lines

delete_list auth-required never probed

Delete one of the person's lists for good. It disappears for them and for readers and cannot be brought back. Refused for Inbox. Only after the person asks to delete that exact list by name; offer archive_list when they may want it back.

mcp-tool

{
  "type": "object",
  "required": [
    "list",
    "confirmTitle"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Which list: slug, exact title, or id. Must match exactly one list or the call fails."
    },
    "confirmTitle": {
      "type": "string",
      "description": "The list's exact title, after the person confirms deleting it. A different title deletes nothing."
    }
  },
  "additionalProperties": false
}

arguments 18 lines

add_item auth-required never probed

Capture one pick with the person's own note. Omit `list` only to deliberately use Inbox. An ambiguous or missing list writes nothing and returns candidates; ask the person, or set `create_if_missing` when they explicitly want a new list. Report the returned details status exactly.

mcp-tool

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "The URL being captured, if any."
    },
    "list": {
      "type": "string",
      "description": "List slug or title. Fuzzy-matched against the curator's lists. Omit to use Inbox."
    },
    "note": {
      "type": "string",
      "description": "The curator's own note on why they like or recommend this pick. Shown to readers as written and never rewritten by AI, so capture it in the curator's voice when they give a reason."
    },
    "title": {
      "type": "string",
      "description": "Short title for the pick (e.g. restaurant name, book title). Required when url is absent."
    },
    "author": {
      "type": "string",
      "description": "Author or creator as the person gives it (books, articles, posts), e.g. from \"Title - Author - why\". Stored as the pick's author and never replaced by added details. Keep it out of the note."
    },
    "idempotency_key": {
      "type": "string",
      "maxLength": 200,
      "description": "Optional retry key. Reusing it with identical input replays the first result."
    },
    "create_if_missing": {
      "type": "boolean",
      "description": "When true and no list matches `list`, create that private list and save there."
    }
  },
  "additionalProperties": false
}

arguments 35 lines

add_items auth-required never probed

Add up to 50 picks to one list in a single call, in order. Unlike add_item this never falls back to the Inbox: `list` must resolve to exactly one of the curator's lists or nothing is written.

mcp-tool

{
  "type": "object",
  "required": [
    "list",
    "items"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Which list: exact slug, handle/slug, or id from a previous result."
    },
    "items": {
      "type": "array",
      "items": {
        "type": "object",
        "properties": {
          "url": {
            "type": "string",
            "description": "The pick URL, if any."
          },
          "note": {
            "type": "string",
            "description": "The curator's own reason for recommending it. Stored verbatim and never AI-rewritten."
          },
          "title": {
            "type": "string",
            "description": "Item title. Required unless a url is given."
          },
          "author": {
            "type": "string",
            "description": "Author or creator as the person gives it (books, articles, posts), e.g. from \"Title - Author - why\". Stored as the pick's author and never replaced by added details. Keep it out of the note."
          }
        },
        "additionalProperties": false
      },
      "maxItems": 50,
      "description": "Up to 50 picks, added in the order given."
    },
    "idempotency_key": {
      "type": "string",
      "maxLength": 200,
      "description": "Stable key for this reviewed batch. Reuse the same key and identical picks after a lost response."
    }
  },
  "additionalProperties": false
}

arguments 46 lines

import_from_url auth-required never probed
Import picks into a list from one link: a Google Maps list shared with "Anyone with the link" (every place, in order, with the owner's notes; pasting it again adds only new places), a Google Maps place (one pick) or a public YouTube playlist (up to 500 videos). Spotify playlists and Letterboxd lists no longer let apps read them; the tool says so, and the person can bring their export in Curated instead. For a whole list you want to sort first, use start_import.

mcp-tool
{ "type": "object", "required": [ "list", "url" ], "properties": { "url": { "type": "string", "description": "A Google Maps shared list or place link (maps.app.goo.gl works), Spotify playlist, YouTube playlist (a link with list=) or Letterboxd list URL." }, "list": { "type": "string", "description": "Which list: exact slug, handle/slug, or id from a previous result." } }, "additionalProperties": false }
arguments 18 lines

start_import auth-required never probed

Bring a whole list in to sort, privately. Pass `url` for a Google Maps list shared with "Anyone with the link" (Curated reads every place), or `picks` you read yourself from anywhere else: a page, an account, a document or an export (up to 1,000 per call; send more with stage_picks). Nothing is added to a list until commit_import, and nothing is published.

mcp-tool

{
  "type": "object",
  "properties": {
    "url": {
      "type": "string",
      "description": "A Google Maps shared-list link (maps.app.goo.gl works). Omit when sending picks."
    },
    "list": {
      "type": "string",
      "description": "Optional destination list. Otherwise commit_import can make a new private list."
    },
    "sort": {
      "type": "boolean",
      "description": "True when most entries are probably not recommendations (an account's posts, saved posts): everything arrives unsorted for the person to keep or skip. Default false: everything arrives kept."
    },
    "picks": {
      "type": "array",
      "items": {
        "type": "object",
        "required": [
          "title"
        ],
        "properties": {
          "url": {
            "type": "string"
          },
          "note": {
            "type": "string",
            "description": "The person's own words, only when the source is theirs."
          },
          "title": {
            "type": "string"
          },
          "author": {
            "type": "string",
            "description": "Author or creator as the person gives it (books, articles, posts), e.g. from \"Title - Author - why\". Stored as the pick's author and never replaced by added details. Keep it out of the note."
          },
          "saved_at": {
            "type": "string",
            "description": "When it was posted or saved at the source, ISO 8601."
          },
          "entry_key": {
            "type": "string",
            "description": "A stable id for this entry in its source (a post's permalink id), so running again brings only new entries."
          },
          "item_type": {
            "type": "string",
            "description": "place, book, movie, music, post, product or another Curated type, when obvious."
          },
          "source_text": {
            "type": "string",
            "description": "Someone else's words about it (a caption or review), quoted exactly and short. Kept credited, never the person's note."
          },
          "source_account": {
            "type": "string",
            "description": "Who wrote source_text, e.g. \"@tinnedfish\"."
          }
        },
        "additionalProperties": false
      },
      "maxItems": 1000
    },
    "title": {
      "type": "string",
      "description": "The source list's own name, e.g. \"@tinnedfish's posts\" or \"Films I rewatch\". Required with picks."
    },
    "own_list": {
      "type": "boolean",
      "description": "True (default) when the source is the person's own: their notes become their notes. False for someone else's list or account: words stay credited to them."
    },
    "source_owner": {
      "type": "string",
      "description": "Whose list or account it is, when it is not the person's own, e.g. \"@tinnedfish\". Credits them on the list."
    },
    "idempotency_key": {
      "type": "string",
      "maxLength": 200,
      "description": "Optional retry key. Reusing it with identical input replays the first result."
    }
  },
  "additionalProperties": false
}

arguments 82 lines

stage_picks auth-required never probed

Add up to 1,000 more picks to an import started with start_import, for sources longer than one call.

mcp-tool

{
  "type": "object",
  "required": [
    "import_id",
    "picks"
  ],
  "properties": {
    "sort": {
      "type": "boolean",
      "description": "Same meaning as in start_import."
    },
    "picks": {
      "type": "array",
      "items": {
        "type": "object",
        "additionalProperties": true
      },
      "maxItems": 1000
    },
    "import_id": {
      "type": "string",
      "description": "The id start_import returned."
    }
  },
  "additionalProperties": false
}

arguments 26 lines

review_import auth-required never probed

Read an import's entries with their ids and states (kept, skipped, unsorted), to show the person or to sort.

mcp-tool

{
  "type": "object",
  "required": [
    "import_id"
  ],
  "properties": {
    "state": {
      "enum": [
        "pending",
        "kept",
        "skipped",
        "committed"
      ],
      "type": "string",
      "description": "Only entries in this state."
    },
    "offset": {
      "type": "integer",
      "minimum": 0
    },
    "import_id": {
      "type": "string"
    }
  },
  "additionalProperties": false
}

arguments 26 lines

sort_import auth-required never probed

Keep or skip entries in an import, by id or all unsorted at once, as the person decides. Never decide for them which posts are recommendations; suggest, then ask.

mcp-tool

{
  "type": "object",
  "required": [
    "import_id"
  ],
  "properties": {
    "keep": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "Entry ids to keep."
    },
    "rest": {
      "enum": [
        "keep",
        "skip"
      ],
      "type": "string",
      "description": "Then keep or skip every entry still unsorted."
    },
    "skip": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "Entry ids to skip."
    },
    "import_id": {
      "type": "string"
    }
  },
  "additionalProperties": false
}

arguments 34 lines

commit_import auth-required never probed

Add the kept entries of an import to a list, privately: the list given at start_import, another list of theirs, or a new private list. Safe to retry; nothing is added twice. A Google Maps list can stay connected so new saves there show up later.

mcp-tool

{
  "type": "object",
  "required": [
    "import_id"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Destination, when start_import had none. Omit to make a new private list."
    },
    "connect": {
      "type": "boolean",
      "description": "Keep a Google Maps list connected (default true)."
    },
    "import_id": {
      "type": "string"
    },
    "new_list_title": {
      "type": "string",
      "description": "Title for a new list. Defaults to the source's title."
    }
  },
  "additionalProperties": false
}

arguments 24 lines

list_sources auth-required never probed

Show what one of the person's lists is connected to: when it was last checked, what is new and waiting, and what is gone at the source.

mcp-tool

{
  "type": "object",
  "required": [
    "list"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Which list: exact slug, handle/slug, or id from a previous result."
    }
  },
  "additionalProperties": false
}

arguments 13 lines

check_source auth-required never probed

Read a connected list's source again now (up to 10 times a day per list). New entries arrive by the list's policy; nothing is removed.

mcp-tool

{
  "type": "object",
  "required": [
    "source_id"
  ],
  "properties": {
    "policy": {
      "enum": [
        "ask",
        "add_private"
      ],
      "type": "string",
      "description": "Optionally change what happens to new entries first: wait to be sorted (ask) or be added privately (add_private). Add and publish is set in Curated, after the person verifies the list is theirs."
    },
    "source_id": {
      "type": "string",
      "description": "A connection id from list_sources."
    }
  },
  "additionalProperties": false
}

arguments 21 lines

list_items auth-required never probed

Read every pick in one of your lists, in reader order, with item ids. Call this before update_item or remove_item to get the id you need.

mcp-tool

{
  "type": "object",
  "required": [
    "list"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Which list: exact slug, handle/slug, or id from a previous result."
    }
  },
  "additionalProperties": false
}

arguments 13 lines

search_items auth-required never probed

Search the curator's own picks by title or URL. Useful to check for a duplicate before add_item.

mcp-tool

{
  "type": "object",
  "required": [
    "query"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Optional list slug or title to scope the search."
    },
    "query": {
      "type": "string",
      "description": "Search string."
    }
  },
  "additionalProperties": false
}

arguments 17 lines

restore_item auth-required never probed
Put a removed pick back on its list where it was, with the same id, note, details, links and saves. Changes the working version only; readers keep seeing the last approved version until the owner approves a list update.

mcp-tool
{ "type": "object", "required": [ "itemId" ], "properties": { "itemId": { "type": "string", "description": "removedId from remove_item, or an id from removed_items." } }, "additionalProperties": false }
arguments 13 lines

update_item auth-required never probed

Edit one pick's title, curator note, URL or author. The note is the curator's own words and is never rewritten by AI, so preserve their voice.

mcp-tool

{
  "type": "object",
  "required": [
    "itemId"
  ],
  "properties": {
    "url": {
      "type": "string",
      "description": "New URL. Pass an empty string to clear it."
    },
    "note": {
      "type": "string",
      "description": "New curator note. Pass an empty string to clear it."
    },
    "title": {
      "type": "string",
      "description": "New title."
    },
    "author": {
      "type": "string",
      "description": "Author or creator as the person gives it (books, articles, posts), e.g. from \"Title - Author - why\". Stored as the pick's author and never replaced by added details. Keep it out of the note. Pass an empty string to remove it."
    },
    "itemId": {
      "type": "string",
      "description": "Item id from list_items or search_items."
    }
  },
  "additionalProperties": false
}

arguments 29 lines

remove_item auth-required never probed
Remove one pick from a list. Curated keeps it with its note, links and saves, and restore_item puts it back. Confirm with the curator first. Changes the working version only; readers keep seeing the last approved version until the owner approves a list update.

mcp-tool
{ "type": "object", "required": [ "itemId" ], "properties": { "itemId": { "type": "string", "description": "Item id from list_items or search_items." } }, "additionalProperties": false }
arguments 13 lines

reorder_items auth-required never probed

Put a list's picks in a new order. Pass every item id exactly once, first to last, and the `revision` from list_items; if the list changed since, nothing is written and the current revision comes back. Changes the working order only; readers keep seeing the last approved version until the owner approves a list update. Private lists stay private.

mcp-tool

{
  "type": "object",
  "required": [
    "list",
    "order",
    "revision"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Which list: exact slug, handle/slug, or id from a previous result."
    },
    "order": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "maxItems": 1000,
      "description": "Every item id in the list, in the new order."
    },
    "revision": {
      "type": "string",
      "description": "The `revision` list_items returned."
    }
  },
  "additionalProperties": false
}

arguments 27 lines

move_items auth-required never probed

Move picks from one of the person's lists to another, keeping each pick's note, details and history. They go to the end of the destination in the order given. Pass the source list's `revision` from list_items; if it changed since, nothing moves. Changes the working versions only; readers keep seeing the last approved versions. On a live destination, call review_items and confirm_items for the moved picks after the person's approval. Updating the source's published version requires separate owner approval. Private lists stay private.

mcp-tool

{
  "type": "object",
  "required": [
    "from",
    "to",
    "itemIds",
    "revision"
  ],
  "properties": {
    "to": {
      "type": "string",
      "description": "The list to move them to."
    },
    "from": {
      "type": "string",
      "description": "The list the picks are in now."
    },
    "itemIds": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "maxItems": 100,
      "description": "Item ids from list_items on the `from` list."
    },
    "revision": {
      "type": "string",
      "description": "The `revision` list_items returned for the `from` list."
    },
    "toRevision": {
      "type": "string",
      "description": "Optional: the destination revision, to refuse the move if it changed too."
    }
  },
  "additionalProperties": false
}

arguments 36 lines

save_list auth-required never probed
Save a live list to Library under Saved lists, with its new picks. Never follows the curator, turns on email or publishes. Idempotent across retries.

mcp-tool
{ "type": "object", "required": [ "list" ], "properties": { "list": { "type": "string", "description": "Exact list id or slug." } }, "additionalProperties": false }
arguments 13 lines

saved_lists auth-required never probed

Read back Saved lists, matching the app. Each one brings its new picks.

mcp-tool

{
  "type": "object",
  "properties": {
    "limit": {
      "type": "number",
      "maximum": 100,
      "minimum": 1
    }
  },
  "additionalProperties": false
}

arguments 11 lines

saved_items auth-required never probed

Read back your saved picks across every list, newest first, matching Saved picks in the app. Separate from saved_lists and from my_library access.

mcp-tool

{
  "type": "object",
  "properties": {
    "list": {
      "type": "string",
      "description": "Optional list id, slug or handle/slug to show saves from one list."
    },
    "limit": {
      "type": "number",
      "maximum": 100,
      "minimum": 1
    },
    "cursor": {
      "type": "string",
      "description": "The cursor from a previous saved_items result, to read further."
    }
  },
  "additionalProperties": false
}

arguments 19 lines

add_saved_item auth-required never probed

Add a pick you can read, usually one from saved_items, to one of your own lists as a new pick. Carries the facts (title, link, place, type) and records where it was found; the note starts empty because the source curator's words are theirs. Does not change the source list or the save.

mcp-tool

{
  "type": "object",
  "required": [
    "itemId",
    "list"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Your list slug or title. Call list_my_lists if unsure."
    },
    "itemId": {
      "type": "string",
      "description": "Item id from saved_items or get_list_items."
    },
    "idempotency_key": {
      "type": "string",
      "description": "Stable across retries so a lost response does not add the pick twice."
    }
  },
  "additionalProperties": false
}

arguments 22 lines

search_library auth-required never probed

Search inside the lists this person has access to: "what was that ramen place from the Tokyo list?". Only searches lists they can read in full.

mcp-tool

{
  "type": "object",
  "required": [
    "query"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Optional: restrict to one list (handle/slug, slug, title, or id)."
    },
    "limit": {
      "type": "number",
      "description": "Max picks, 1 to 50. Defaults to 20."
    },
    "query": {
      "type": "string",
      "description": "Search string, matched against pick titles, URLs, and descriptions."
    }
  },
  "additionalProperties": false
}

arguments 21 lines

library_updates auth-required never probed

What's new in the lists this person has access to since a point in time. Defaults to the last 7 days. Pass the returned `checkedAt` back as `since` next time for a gap-free diff.

mcp-tool

{
  "type": "object",
  "properties": {
    "limit": {
      "type": "number",
      "description": "Max new picks, 1 to 200. Defaults to 50."
    },
    "since": {
      "type": "string",
      "description": "ISO 8601 timestamp. Defaults to 7 days ago."
    }
  },
  "additionalProperties": false
}

arguments 14 lines

review_items auth-required never probed
After adding or importing picks, fetch their enriched preview here. If preparing, wait briefly and call again. Render previewMarkdown: supplied artwork, title/byline, original note, short description and labeled links for every pick. Keep reviewId and itemRefs internal. Omit IDs, positions, confidence scores and provenance commentary. Ask once if it looks good. No app visit or later list update is needed. An unfinished preview survives reconnecting.

mcp-tool
{ "type": "object", "required": [ "list" ], "properties": { "list": { "type": "string", "description": "Which list: exact slug, handle/slug, or id from a previous result." } }, "additionalProperties": false }
arguments 13 lines
pick_review_status auth-required never probed
Read whether an exact pick preview was already confirmed, is pending, or expired. Use when reopening a previous preview or checking a lost confirmation response. Never adds picks, starts enrichment, or changes the review. Keep reviewId internal. If confirmed, do not ask for approval again.

mcp-tool
{ "type": "object", "required": [ "reviewId" ], "properties": { "reviewId": { "type": "string", "description": "reviewId from review_items." } }, "additionalProperties": false }
arguments 13 lines
confirm_items auth-required never probed
After the person approves the enriched preview from review_items, finish adding those exact picks. Published lists show them immediately; private lists stay private. Never approves unrelated edits. Requires lists.publish only for a published destination. If the preview changed, call review_items and show the refreshed picks.

mcp-tool
{ "type": "object", "required": [ "reviewId" ], "properties": { "reviewId": { "type": "string", "description": "reviewId from review_items." } }, "additionalProperties": false }
arguments 13 lines
publish_list auth-required never probed
Prepare first publication or a list-wide update. For new picks use review_items and confirm_items during addition. Publishes nothing. Returns exactly what would become public and, when this connection may publish from chat, a review to confirm with confirm_publication after the owner agrees. Otherwise it returns a browser link where the owner publishes. Show the supplied public URL as a clickable link labeled "Link after publishing", distinct from the owner-only review link. The returned preview is the exact public content: present it as written, with no added images, layouts or caveats.

mcp-tool
{ "type": "object", "required": [ "list" ], "properties": { "list": { "type": "string", "description": "Which list: exact slug, handle/slug, or id from a previous result." } }, "additionalProperties": false }
arguments 13 lines

confirm_publication auth-required never probed

Publish the exact list the owner reviewed, after they explicitly agree to that summary in this conversation. Never on an earlier general instruction like 'publish when done'. If anything changed since the review, nothing is published and a fresh review comes back to show them.

mcp-tool

{
  "type": "object",
  "required": [
    "reviewId"
  ],
  "properties": {
    "handle": {
      "type": "string",
      "description": "Only when the review showed an unconfirmed handle: the same handle, after the owner confirms it's theirs to show publicly."
    },
    "reviewId": {
      "type": "string",
      "description": "review.reviewId from publish_list."
    }
  },
  "additionalProperties": false
}

arguments 17 lines

search_lists auth-required never probed

Search every live list on Curated, not just the person's own. Each result carries `access` (yours, unlocked or locked) and a `ref` of the form handle/slug, the unambiguous way to name a list in get_list.

mcp-tool

{
  "type": "object",
  "required": [
    "query"
  ],
  "properties": {
    "limit": {
      "type": "number",
      "description": "Max results, 1 to 25. Defaults to 10."
    },
    "query": {
      "type": "string",
      "description": "What to look for, e.g. \"natural wine bars in Lisbon\"."
    },
    "category": {
      "type": "string",
      "description": "Optional category slug to narrow results, e.g. \"restaurants\"."
    }
  },
  "additionalProperties": false
}

arguments 21 lines

find_lists auth-required never probed

Search live Curated lists by topic, title, or curator handle. Returns exact slugs; use get_list_items to read one.

mcp-tool

{
  "type": "object",
  "required": [
    "query"
  ],
  "properties": {
    "query": {
      "type": "string",
      "description": "What the user is looking for."
    }
  },
  "additionalProperties": false
}

arguments 13 lines

get_list auth-required never probed

Read one list: every pick of a live list, or of one of the person's own. Accepts handle/slug, a bare slug, a title, or an id.

mcp-tool

{
  "type": "object",
  "required": [
    "list"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Which list: handle/slug, slug, title, or id."
    },
    "format": {
      "enum": [
        "text",
        "markdown"
      ],
      "type": "string",
      "description": "How to render the reply text. 'text' (default): numbered rows, one field per line, best for reading the list yourself. 'markdown': linked bold titles and quoted notes, for relaying the list to the user as-is."
    }
  },
  "additionalProperties": false
}

arguments 21 lines

get_list_items auth-required never probed

Read one list and report each pick's details status. The capture read-back check; use the exact slug returned by create_list or list_my_lists.

mcp-tool

{
  "type": "object",
  "required": [
    "list"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Exact returned slug, handle/slug, title, or id."
    },
    "format": {
      "enum": [
        "text",
        "markdown"
      ],
      "type": "string",
      "description": "How to render the reply text. 'text' (default): numbered rows, one field per line, best for reading the list yourself. 'markdown': linked bold titles and quoted notes, for relaying the list to the user as-is."
    }
  },
  "additionalProperties": false
}

arguments 21 lines

save_item auth-required never probed

Save one pick from a list you can read, yours or anyone's, to Library under Saved picks. Use the item id from get_list_items or list_items. Does not copy the pick, follow or publish. Idempotent across retries.

mcp-tool

{
  "type": "object",
  "required": [
    "itemId"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Optional list id, slug or handle/slug the pick came from; rejected if the pick is on a different list."
    },
    "itemId": {
      "type": "string",
      "description": "Item id from get_list_items, list_items or search_items."
    }
  },
  "additionalProperties": false
}

arguments 17 lines

unsave_item auth-required never probed

Remove one pick from Saved picks. Idempotent. Never changes the list or the pick.

mcp-tool

{
  "type": "object",
  "required": [
    "itemId"
  ],
  "properties": {
    "itemId": {
      "type": "string",
      "description": "Item id from saved_items."
    }
  },
  "additionalProperties": false
}

arguments 13 lines

get_list_updates auth-required never probed
Older name for save_list on a live free list: saves it to Library with its new picks. Only when asked. Idempotent; no curator follow, email or payment. Read back with saved_lists.

mcp-tool
{ "type": "object", "required": [ "list" ], "properties": { "list": { "type": "string" } }, "additionalProperties": false }
arguments 12 lines

follow_curator auth-required never probed

Explicitly follow or unfollow a curator by handle. Idempotent desired state, never a toggle. Does not save a list or enable email.

mcp-tool

{
  "type": "object",
  "required": [
    "handle",
    "following"
  ],
  "properties": {
    "handle": {
      "type": "string"
    },
    "following": {
      "type": "boolean"
    }
  },
  "additionalProperties": false
}

arguments 16 lines

curator_notifications auth-required never probed
Read a followed curator's email preference, or set `enabled` only on an explicit email request. Follow is separate consent; list access and bookmarks are unchanged.

mcp-tool
{ "type": "object", "required": [ "handle" ], "properties": { "handle": { "type": "string" }, "enabled": { "type": "boolean" } }, "additionalProperties": false }
arguments 15 lines
list_notifications auth-required never probed
Read every notification setting when `push` and `email` are omitted. Only on an explicit request, set push and/or email about new picks on one list the person gets updates for. Never changes access, saves or follows.

mcp-tool
{ "type": "object", "properties": { "list": { "type": "string", "description": "List id, slug or handle/slug; required when setting." }, "push": { "type": "boolean" }, "email": { "type": "boolean" } }, "additionalProperties": false }
arguments 16 lines

unlock_list auth-required never probed

Save a live free list with its new picks, only when the person asks. Same result as save_list. Idempotent; never charges anyone.

mcp-tool

{
  "type": "object",
  "required": [
    "list"
  ],
  "properties": {
    "list": {
      "type": "string",
      "description": "Which list: handle/slug, slug, title, or id."
    }
  },
  "additionalProperties": false
}

arguments 13 lines

account_context auth-required 1h ago
Verify the connected account and its readiness before creating or changing lists. Returns account requirements, recovery URLs and what this connection may do. Never ask for a token in chat.

mcp-tool
{ "type": "object", "properties": {}, "additionalProperties": false }
arguments 5 lines
my_library auth-required 1h ago
The lists this person saved, each bringing its new picks. Their own lists are not included; use list_my_lists for those.

mcp-tool
{ "type": "object", "properties": { "limit": { "type": "number", "description": "Max lists, 1 to 200. Defaults to 50." } }, "additionalProperties": false }
arguments 10 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.

Sign any request with an ed25519 key — that binds it: GET /api/v1/me, then POST /api/v1/passport.
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.
Ask the hub to check: POST /api/v1/passport/claim-endpoint with this listing's id 85470c9c90525ca1.

Every step, filled in for this listing: https://brick.blue/api/v1/agents/85470c9c90525ca1/claim. Over MCP: the claim_endpoint tool.

_ for your README measured, not declared

[![measured by brick.blue](https://brick.blue/api/v1/agents/85470c9c90525ca1/badge.svg)](https://brick.blue/agent/85470c9c90525ca1?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

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.