Skip to content

ZenVault MCP server

Which MCP servers are ZenVault'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 ZenVault

Whose server it is
  • Registry nameeu.zenvault/zenvaultverified in the MCP registry

The registry name sits under zenvault.eu. 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
  • Tools44
  • Server calls itselfzenvault
  • Server version1.0.0
  • Protocol version2024-11-05

On 2026-10-26 the server answered initialize and tools/list without a sign-in. First answer in 810 ms.

Should an agent connect itNot proven either way yet
  • Its tools do every use case we checked: schedule webhook wake-ups for recurring agent tasks; watch web pages and alert when content changes; buffer inbound webhooks until the agent collects them.
  • Checked on 1 of the 7 days it takes to judge reliability.

Once connected

  • For "schedule webhook wake-ups for recurring agent tasks", the likely first call is create_schedule.
  • For "watch web pages and alert when content changes", the likely first call is create_watch.
  • For "buffer inbound webhooks until the agent collects them", the likely first call is create_inbox.
  • create_wakeup, create_schedule, create_deadman, create_drip, create_inbox change data without a sign-in: anyone with the address can call them.
  • Parameters are untyped or undescribed in 33 of 44 tools: check arguments before calling.

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
  • Answeredmeasuring1 of the 7 daily checks it takes to print a share
  • Median first answer859 ms
  • Space in an agent's context≈ 3,231 tokensthe definitions of all its tools (44), which an agent carries on every turn
  • Tool definitionsC · 71 of 100
  • Typed results0 of 44 toolsdeclared outputSchema: results come typed, read once
  • Parameters without a description: in 33 of 44
  • Not saying whether it only reads or can destroy: 44 of 44

What an agent can do with it

  • can do · schedule webhook wake-ups for recurring agent tasks
  • can do · watch web pages and alert when content changes
  • can do · buffer inbound webhooks until the agent collects them

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: create_wakeup, create_schedule, create_deadman, create_drip, create_inbox

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 offers44 tools · read 2026-10-26
  • create_wakeup

    Schedule a webhook call at a future time (one-shot). Your agent can die; the call still happens.

  • create_schedule

    Recurring webhook on a cron expression until the prepaid days run out.

  • create_deadman

    Dead-man switch: if you stop checking in, we fire your alert webhook.

  • checkin_deadman

    Check in to a dead-man switch so it does not fire.

  • delayed_delivery

    Hand over a payload now; we POST it to the target at a future moment.

  • create_drip

    Dump up to 100 items now; we deliver them to your webhook at a fixed rate.

  • notarize

    Signed proof that a hash (or your data) existed at this moment.

  • attested_fetch

    Fetch a URL and return content plus a signed receipt binding url, status, content hash and time.

  • create_inbox

    Public URL that buffers inbound webhooks until you collect them.

  • collect_messages

    Collect buffered messages from an inbox, bin, front door or form.

  • create_frontdoor

    Durable front door: buffers webhooks and forwards them to your gateway with retry-until-ack.

  • request_approval

    Ask a human to approve or deny an action; returns a URL for them and a signed verdict for you. Times out as denied.

  • check_approval

    Check whether a human approved or denied your request.

  • notify_human

    Send a message to a human on Telegram, Discord, Slack, ntfy, email and 100+ other services (apprise URL is used once, never stored).

  • guaranteed_delivery

    POST a payload to a URL, retried with backoff until it succeeds.

  • create_vault

    Create a key-value namespace with TTL that survives your runs.

  • kv_set

    Store a value in a vault (max 1 MB).

  • kv_get

    Read a value from a vault.

  • kv_list

    List keys in a vault.

  • save_checkpoint

    Save run state so you can resume after a crash.

  • create_checkpoint

    Create a crash checkpoint store.

  • load_checkpoint

    Load the latest saved checkpoint state.

  • store_secret

    Store a value that self-destructs on first read; returns a one-time read URL.

  • create_queue

    Pull-based work queue with leases; timed-out leases requeue.

  • queue_push

    Push a job onto a queue.

  • queue_lease

    Lease the next job from a queue.

  • queue_ack

    Acknowledge (or nack) a leased job.

  • create_counter

    Create an atomic counter.

  • counter_incr

    Increment a counter atomically.

  • claim_operation

    Idempotency firewall: claim an operation id exactly once across runs.

  • create_idempotency

    Create an idempotency/replay firewall.

  • publish_page

    Publish an HTML page at name.zenvault.eu for N days.

  • create_link

    Create an expiring share link with click stats.

  • create_watch

    Watch something and fire your webhook when it changes. Types: change, keyword, regex, structured, drift, redirect, sitemap, staleness, rss, dns, port, cert, heartbeat.

  • watch_status

    Check a watch: last value, checks, fires.

  • create_escalation

    Escalation chain: webhook, then after N seconds the next step, until acknowledged.

  • get_resource

    Status of any resource you hold the management token for.

  • extend_resource

    Extend a resource TTL by paying again.

  • cancel_resource

    Cancel a resource early.

  • export_resource

    Export everything under a resource before it expires.

  • credit_balance

    Check the balance of your prepaid credit token.

  • create_subtoken

    Mint a spend-capped child token for a sub-agent.

  • list_prices

    Machine-readable price list for every ZenVault service, in USD.

  • request_feature

    Tell the operator what ZenVault is missing. Free.

How to connect

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

{
  "mcpServers": {
    "zenvault": {
      "type": "http",
      "url": "https://api.zenvault.eu/mcp"
    }
  }
}

Add it

Claude Code, in a terminal

claude mcp add --transport http zenvault https://api.zenvault.eu/mcp

Add to CursorAdd to VS Code

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

OpenAPI spec

52 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.