Skip to content
Avatar 8 LLC logo

Avatar 8 LLC MCP server

Which MCP servers are Avatar 8 LLC's own, from the official registry and from asking its own domain, whether they answer, and what they offer. A program checked each of these, nobody copied them from a product page.

read 2026-10-26

← Back to Avatar 8 LLC

Whose server it is
  • Registry namepage.deed/deed-pageverified in the MCP registry

The registry name sits under deed.page. The registry only lets a publisher use that namespace after proving control of the domain through DNS or HTTP.

When we asked itchecked 2026-10-26
  • Stateanswerschecked 2026-10-26
  • Tools26
  • Server calls itselfdeed.page
  • Server version1.9.9
  • Protocol version2025-06-18

On 2026-10-26 the server answered initialize and tools/list without a sign-in. First answer in 177 ms.

Should an agent connect itNot proven either way yet
  • Its tools do every use case we checked: publish generated files to a permanent public URL; update a live site under the same slug; add a custom domain to a published site.
  • Checked on 1 of the 7 days it takes to judge reliability.

Once connected

  • For "publish generated files to a permanent public URL", the likely first call is publish_site.
  • For "update a live site under the same slug", the likely first call is update_site.
  • For "add a custom domain to a published site", the likely first call is add_domain.
  • publish_site, update_site, delete_site, add_domain, remove_domain change data without a sign-in: anyone with the address can call them.

Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.

Measured dailydaily checks: 1, the first on 2026-10-26
  • INITE MCP scoreafter a weeka score needs 7 daily checks: 1 so far
  • What an agent can do with it3 of 3 use cases
  • Use2 GitHub stars
  • Answeredmeasuring1 of the 7 daily checks it takes to print a share
  • Median first answer211 ms
  • Space in an agent's context≈ 6,834 tokensthe definitions of all its tools (26), which an agent carries on every turn
  • Tool definitionsA · 98 of 100
  • Typed results0 of 26 toolsdeclared outputSchema: results come typed, read once
  • Parameters without a description: in 3 of 26

What an agent can do with it

  • can do · publish generated files to a permanent public URL
  • can do · update a live site under the same slug
  • can do · add a custom domain to a published site

Each of the product's use cases, put to Jev against this server's tool list: can an agent do it with these tools?

Worth a look

  • Tools that change data, on a server that answers without a sign-in: publish_site, update_site, delete_site, add_domain, remove_domain

A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (25%), doing the product's job (25%), definitions (20%), context (10%), use (10%) and how readily a client connects (10%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →

Tools it offers26 tools · read 2026-10-26
  • publish_site

    Publishes files to a live permanent URL at a slug. Inputs: files (required), and optional slug, spa, ttl_seconds, merge, expected_version, via, heard_from, token, listed, category, tags, title, and description. Paths are stored exactly…

  • update_site

    Replaces the files of a slug owned by the token, or, when merge is true, adds and overwrites only the paths sent. Inputs: slug, files, and optional spa, merge, expected_version, via, token, listed, category, tags, title, and description…

  • get_site

    Returns whether a slug is live, plus its URL, version, size, and timestamps. Input: slug. No token is required.

  • list_sites

    Returns every live site owned by the token, with URLs and versions. Input: token, or an Authorization bearer header.

  • delete_site

    Permanently removes a slug owned by the token and detaches its custom hostnames. The slug becomes available again. Inputs: slug and token.

  • rotate_token

    Issues a new token for the same principal. Every site carries over. Inputs: token, and optional overlap_seconds from 0 to 300. The previous token stops working immediately when overlap_seconds is omitted or 0. During a non-zero overlap…

  • revoke_token

    Revokes owner tokens. Inputs: token, and optional scope. Scope "current" (the default) revokes the token that was sent. Scope "all" revokes every token for that principal, including any overlap window. A later request with a revoked token…

  • whoami

    Returns the plan (plan.id is free, hobby, or pro), limits, the number of sites the token owns, and the number of custom hostnames attached. Use it to confirm a plan change after payment. Input: token.

  • get_upgrade_link

    Creates a Stripe Checkout URL that upgrades this token's owner to Hobby ($3/mo) or Pro ($15/mo). Call it when the user asks to upgrade, or right after a tool error that has upgrade_url / user_message (a plan limit). Show the returned…

  • claim_share_credit

    Free plan only, once per token. If the user posted publicly about deed.page on X or Reddit, pass that post's URL. The server checks that the post is public, mentions deed.page or links to a deed.page site, is from the last 30 days, and…

  • list_versions

    Returns every retained version of a slug owned by the token, newest first. Inputs: slug and token. Each entry includes version, created_at, file_count, bytes, kind (publish, merge, or restore), restored_from, and a public permalink. Free…

  • restore_version

    Creates a new version whose files equal a retained version, without uploading those files again. Inputs: slug, version, optional expected_version, and token. When expected_version does not match the live version, the result is…

  • add_domain

    Attaches a hostname to a slug owned by the token and returns the DNS records to create: a CNAME for a subdomain, an A record or an ALIAS record for an apex (one of those, not both), and every TXT record. Inputs: slug, domain, and token…

  • list_domains

    Returns custom hostnames on one slug, or on every site the token owns when slug is omitted. Inputs: token, and optional slug. Status is the last checked value.

  • remove_domain

    Removes a hostname so requests to that host no longer serve the site. Inputs: domain, optional slug, and token. When slug is set, the hostname must already be attached to that slug.

  • domain_status

    Returns the verification status of a custom domain and refreshes verification, DNS config, and certificate state. Inputs: slug, domain, and token. status is pending_dns, verifying, issuing_cert, active, or error. dns_records are the…

  • plan_upload

    Accepts a file manifest and returns which hashes are already stored, a direct upload URL for each missing hash, and a session id. Inputs: files as {path, sha256, size, content_type?}, and optional slug, spa, merge, ttl_seconds…

  • finalize_upload

    Checks uploaded hashes, screens the new bytes, and commits one version under the slug lock. Inputs: session (upl_...), optional expected_version, via, heard_from, and token. A stale expected_version returns version_conflict and writes…

  • get_manifest

    Returns the canonical manifest for a site version: paths, sha256, sizes, version, slug, and created_at, plus an Ed25519 signature. signature is null when the server has no signing key. Inputs: slug, and optional version. When version is…

  • site_analytics

    Returns owner-only cookieless counts for a slug: views per UTC day, top paths, top referrer hosts, and AI crawler counts by user-agent family. Inputs: slug, optional from and to as YYYY-MM-DD (default the last 7 days, maximum 90), and…

  • set_site_access

    Changes the access setting stored for a slug on deed.page. Inputs: slug, password, and token. password is 8 to 200 characters, or an empty string (or null) to clear the password. The call writes a scrypt hash into the site record on…

  • set_listing

    Sets or clears the public gallery listing for a slug. Inputs: slug, listed, and when listed is true a category, title, description, and optional tags (at most 5), plus token. Returns listed, status, category, tags, title, and description…

  • set_badge

    Free-plan sites show a small 'Published with deed.page' link in the bottom-right corner; Hobby and Pro never show it. Inputs: show (boolean), optional slug, and token. With slug, the setting applies to that site; without slug, it applies…

  • create_webhook

    Registers an http or https URL for the events site.published, site.version_restored, site.deleted, site.taken_down, domain.active, and report.received. Inputs: url, optional events, and token. Returns the webhook id and an HMAC secret…

  • list_webhooks

    Returns the webhook URLs registered for the token. Secrets are not included. Input: token.

  • delete_webhook

    Deletes one webhook. Inputs: id (whk_...) and token.

How to connect

Add the server to your MCP client config. Claude Desktop, Cursor and most other clients read this shape:

{
  "mcpServers": {
    "deed-page": {
      "type": "http",
      "url": "https://deed.page/mcp"
    }
  }
}

Add it

Claude Code, in a terminal

claude mcp add --transport http deed https://deed.page/mcp

Add to CursorAdd to VS Code

claude.ai, ChatGPT and other clients: add https://deed.page/mcp as a custom connector in the settings.

OpenAPI spec

48 paths, read 2026-10-26. Open the spec ↗

Each line is something a program found when it went looking: the registry entry, the server's own answer, a spec that parses. A missing line means we did not find it, not that it does not exist.