experimental-ext-triggers-events/docs/design-sketch-proposal.md

MCP Events — Design Sketch

Status: Draft proposal Authors: Peter Alexander Date: 2026-02-19

Summary

Add an Events primitive to MCP that allows clients to subscribe to things happening in the outside world (new email, Slack message, devops alert, scheduled reminder, etc.) and react to them. The design supports three delivery mechanisms — client-side polling, server-side push over an existing connection, and webhook delivery to an external endpoint — with a unified subscription API that abstracts the difference from application code.

Key design principles:

• Servers advertise delivery modes; no mode is mandatory. Each event type lists the delivery modes it supports. The client picks the best mode it can use. If there is no overlap between what the server offers and what the client can consume, that event type is simply unavailable to that client — the protocol does not require a universal fallback. (Reference SDKs are still encouraged to make poll cheap to support so that overlap is common in practice.)
• SDK-level polling, not LLM-level polling. The client SDK drives the polling loop without burning LLM inference tokens. The LLM is only invoked when events actually arrive.
• No durable subscription state. Poll holds no protocol-required subscription state — each request is self-describing. Push scopes subscription state to connection lifetime. Webhook uses soft state with mandatory TTL — the server holds subscriptions in memory, but they expire automatically if the client stops refreshing. The client is always the source of truth across all three modes. (This principle is about protocol-required state. SDKs MAY hold ephemeral derived state — a poll-lease table for lifecycle hooks, an emit ring buffer — but it is reconstructable from subsequent client requests, never persisted, and never owed to any particular client. Servers that relay from upstream sources will also hold upstream credentials, upstream webhook registrations, etc., all outside the protocol's concern.)
• Client owns subscription state. In all modes, the client holds the canonical list of subscriptions. For poll and push, the server holds no protocol-required subscription state. For webhook, the server holds TTL-scoped subscription records, but the client drives their lifecycle via periodic refresh.
• Event payloads are untrusted data. The spec must be explicit that event payloads carry the same injection risks as tool results.
• Servers own delivery. Unlike traditional pub/sub (Kafka, SNS) where a shared broker decouples producers from subscribers, MCP servers run as independent processes with no central infrastructure. Each server handles event delivery itself, with the SDK absorbing the delivery complexity. In the common case, the server is a relay — ingesting events from an upstream system (GitHub webhooks, a Slack socket, a message bus) and routing them to subscribed clients — rather than the originating producer. Deployments that want a decoupled model can use a dedicated MCP server as a broker — no protocol changes required.

Capability Declaration

Servers advertise event support in their capabilities:

{
  "capabilities": {
    "events": {
      "listChanged": true
    }
  }
}

Listing Available Events

Request: events/list

// params (optional)
{ "cursor": "..." }  // pagination

Response

{
  "events": [
    {
      "name": "email.received",
      "description": "Fires when a new email arrives in the inbox",
      "delivery": ["poll"],
      "inputSchema": {
        "type": "object",
        "properties": {
          "from": { "type": "string", "description": "Glob pattern for sender address" },
          "subject_contains": { "type": "string" },
          "redact_pii": { "type": "boolean", "default": false, "description": "Strip PII from event payloads" },
          "include_body_preview": { "type": "boolean", "default": true, "description": "Include a snippet of the email body" }
        }
      },
      "payloadSchema": {
        "type": "object",
        "properties": {
          "messageId": { "type": "string" },
          "from": { "type": "string" },
          "subject": { "type": "string" },
          "receivedAt": { "type": "string", "format": "date-time" }
        }
      }
    },
    {
      "name": "incident.created",
      "description": "Fires when a new PagerDuty incident is created",
      "delivery": ["webhook", "push", "poll"],
      "inputSchema": {
        "type": "object",
        "properties": {
          "severity": { "type": "string", "enum": ["P1", "P2", "P3", "P4"] },
          "service": { "type": "string" },
          "deduplicate_window_seconds": { "type": "integer", "default": 0, "description": "Suppress duplicate alerts within this window" }
        }
      },
      "payloadSchema": { "..." : "..." }
    }
  ]
}

Notes:

• delivery lists the delivery modes this event type supports — any non-empty subset of "poll", "push", "webhook". No mode is mandatory. A client that cannot use any of the listed modes cannot subscribe to this event type.
• inputSchema is a JSON Schema describing valid subscription parameters — these may include filters (which narrow the event stream), transforms (which modify payloads), or other server-defined configuration. This mirrors the inputSchema on tools for consistency.
• payloadSchema describes the shape of data in delivered events.

Dynamic Event Types: notifications/events/list_changed

If the set of available event types changes at runtime (e.g., a plugin is loaded, a data source is connected), the server sends a notifications/events/list_changed notification. The client SHOULD re-call events/list to refresh its event type registry. This is consistent with notifications/tools/list_changed and notifications/resources/list_changed.

Subscribing and Event Delivery

There are three delivery modes with different subscription mechanisms:

• Poll mode: Client calls events/poll with event name, params, and cursor. No separate subscribe step needed — the first poll with a null cursor bootstraps the subscription. Server holds no protocol-required state (the SDK MAY track an ephemeral poll lease for lifecycle hooks; see Unsubscribe timing by mode).
• Push mode: Client opens a long-lived POST (events/stream) carrying all desired subscriptions. Events are delivered on the SSE response stream (HTTP) or as notifications on stdout (stdio). Connection close terminates all subscriptions. Server state is scoped to connection lifetime.
• Webhook mode: Client calls events/subscribe to register a callback URL. The server POSTs events to that URL as they occur. Subscriptions have a mandatory TTL — the client must periodically refresh by re-calling events/subscribe before the TTL expires. If the client stops refreshing, the subscription expires and the server reclaims resources. Designed for remote servers where maintaining a long-lived connection is impractical.

Error Codes

Code	Message	Meaning
-32602	InvalidParams	Params don't match inputSchema (standard JSON-RPC invalid params)
-32011	EventNotFound	Unknown event name
-32012	Unauthorized	User lacks permission for this event/params combination
-32013	TooManySubscriptions	Server-imposed subscription limit reached
-32014	CursorExpired	Cursor is no longer valid (upstream compacted, server reset); client must re-subscribe with cursor: null
-32015	InvalidCallbackUrl	Webhook URL is unreachable or rejected by the server (webhook mode only)
-32016	SubscriptionNotFound	Unknown subscription ID for events/unsubscribe (webhook mode only)

These codes occupy -32011..-32016 to avoid collision with base MCP error codes (e.g., -32002 ResourceNotFound).

Poll-Based Delivery

The client SDK calls events/poll at the server-recommended interval. This is a protocol-level operation, NOT an LLM tool call.

sequenceDiagram
    autonumber
    participant SDK as Client SDK
    participant Server as MCP Server

    loop every nextPollSeconds
        SDK->>Server: events/poll {subscriptions[], cursors}
        Server-->>SDK: {results: [{events[], cursor, nextPollSeconds}]}
    end
    Note over SDK: LLM invoked only when<br/>events[] is non-empty

Request: events/poll

{
  "maxEvents": 50,                    // optional; cap events per subscription
  "subscriptions": [
    {
      "id": "sub_email",              // client-provided identifier
      "name": "email.received",
      "params": {
        "from": "*@anthropic.com",
        "redact_pii": true
      },
      "cursor": null                  // null = start from now
    },
    {
      "id": "sub_incidents",
      "name": "incident.created",
      "params": { "severity": "P1" },
      "cursor": "cursor_abc"          // resume from previous position
    }
  ]
}

Response

{
  "results": [
    {
      "id": "sub_email",
      "events": [
        {
          "eventId": "evt_001",
          "name": "email.received",
          "timestamp": "2026-02-19T15:30:00Z",
          "data": {
            "messageId": "msg_xyz",
            "from": "dsp@anthropic.com",
            "subject": "MCP spec review",
            "receivedAt": "2026-02-19T15:29:58Z"
          }
        }
      ],
      "cursor": "historyId_99842",
      "hasMore": false,
      "nextPollSeconds": 30
    },
    {
      "id": "sub_incidents",
      "events": [],
      "cursor": "cursor_abc",
      "hasMore": false,
      "nextPollSeconds": 60
    }
  ]
}

Notes:

• id is a client-provided identifier for each subscription. It is opaque to the server and echoed back in responses to allow the client to correlate results with subscriptions. It must be unique within a single events/poll request.
• cursor is opaque to the client. The client stores it and passes it back on the next poll. A null cursor means "start from now" — the server returns no events and provides a fresh cursor for subsequent polls.
• eventId enables client-side deduplication across polls (e.g., after a crash/restart). It is server-assigned: when the upstream source provides a stable event identifier (Stripe evt_*, GitHub delivery GUID, Kafka offset, Gmail message ID), the server SHOULD use that value as eventId so that the same upstream event surfaced via multiple paths (e.g., webhook emit and poll backfill) carries the same eventId and dedup works. The SDK auto-generates an eventId only when the author supplies none.
• maxEvents is an optional top-level field that caps the number of events returned per subscription. If more events are available than the limit, the server returns a partial batch with an intermediate cursor and sets hasMore: true. The client SHOULD poll again immediately (ignoring nextPollSeconds) to drain the backlog. If omitted, the server uses its own default limit.
• hasMore indicates whether additional events are available beyond the returned batch. When true, the client should poll again immediately with the updated cursor. When false, the client waits nextPollSeconds before the next poll.
• nextPollSeconds allows the server to dynamically adjust polling frequency per subscription (e.g., back off when rate-limited upstream, speed up when activity is detected). Ignored when hasMore is true.
• Empty events array means nothing happened — this is the common case and should be cheap.
• The server holds no protocol-required per-client subscription state. Each poll request is self-contained: the client provides the event name, params, and cursor. The server does not need to "remember" previous poll requests to answer them. (The SDK MAY hold ephemeral derived state — a poll-lease table to drive on_subscribe/on_unsubscribe, and for emit-only event types a ring buffer of recent events — but neither is required to answer a poll, both are reconstructable, and neither is owed to any particular client. See Unsubscribe timing by mode and Emit-only event types under Server SDK Guidance.)

Error Handling

If an individual subscription within a poll request is invalid, the server returns an error for that subscription without failing the entire request:

{
  "results": [
    {
      "id": "sub_email",
      "events": [ "..." ],
      "cursor": "historyId_99842",
      "nextPollSeconds": 30
    },
    {
      "id": "sub_bogus",
      "error": {
        "code": -32011,
        "message": "EventNotFound"
      }
    }
  ]
}

Push-Based Delivery

Push delivery uses a long-lived events/stream request. The client sends all desired subscriptions, and the server delivers events as they occur. The request is a standard JSON-RPC request with an id, which enables cancellation via notifications/cancelled.

The transport mechanism differs by transport type:

• Streamable HTTP: The events/stream request is a POST that returns an SSE response stream. This stream carries event notifications (notifications/events/*); it is independent of, and does not replace, the transport's existing GET-based SSE stream, which continues to carry non-event server-initiated notifications (notifications/tools/list_changed, progress, logging, etc.). The client cancels by aborting the request stream (TCP close on HTTP/1.1, RST_STREAM on HTTP/2) — no explicit cancellation message is needed.
• stdio: The events/stream request is sent on stdin. The server delivers events as JSON-RPC notifications on stdout. Since there is no connection to close, the client cancels by sending notifications/cancelled with the request's id.

sequenceDiagram
    autonumber
    participant SDK as Client SDK
    participant Server as MCP Server

    SDK->>Server: events/stream {id, subscriptions[]}
    activate Server
    Server-->>SDK: notifications/events/active {id, cursor} ×N
    loop as events occur
        Server-->>SDK: notifications/events/event {id, event, cursor}
    end
    SDK->>Server: notifications/cancelled {requestId} (stdio)<br/>— or connection close (Streamable HTTP)
    Server-->>SDK: StreamEventsResult (final frame)
    deactivate Server

Request: events/stream

// Streamable HTTP: POST /mcp
// stdio: written to stdin
{
  "jsonrpc": "2.0",
  "method": "events/stream",
  "id": 1,
  "params": {
    "subscriptions": [
      {
        "id": "sub_email",
        "name": "email.received",
        "params": { "from": "*@anthropic.com", "redact_pii": true },
        "cursor": null
      },
      {
        "id": "sub_incidents",
        "name": "incident.created",
        "params": { "severity": "P1" },
        "cursor": "cursor_abc"
      }
    ]
  }
}

Event Delivery

The server confirms each subscription, reports errors for invalid ones, and then delivers events as notifications. When the stream ends, the server sends a StreamEventsResult as the final frame:

// Confirmation (one per valid subscription)
{"jsonrpc":"2.0","method":"notifications/events/active","params":{"id":"sub_email","cursor":"historyId_99840"}}
{"jsonrpc":"2.0","method":"notifications/events/active","params":{"id":"sub_incidents","cursor":"cursor_abc"}}

// Error for invalid subscription (stream remains open for valid ones)
{"jsonrpc":"2.0","method":"notifications/events/error","params":{"id":"sub_bogus","error":{"code":-32011,"message":"EventNotFound"}}}

// Events as they occur
{"jsonrpc":"2.0","method":"notifications/events/event","params":{"id":"sub_email","eventId":"evt_001","name":"email.received","timestamp":"2026-02-19T15:30:00Z","data":{"messageId":"msg_xyz","from":"dsp@anthropic.com","subject":"MCP spec review"},"cursor":"historyId_99842"}}

// Final frame when stream closes (StreamEventsResult)
{"jsonrpc":"2.0","id":1,"result":{"_meta":{}}}

On Streamable HTTP, notifications are SSE data: frames; when the server terminates the stream it sends the StreamEventsResult as the final data: frame, but when the client terminates by aborting the request stream no result can be sent (see Cancellation). On stdio, notifications are newline-delimited JSON messages on stdout, and the result is sent when the stream ends.

The events/stream response carries only notifications/events/* messages. Non-event MCP notifications (notifications/tools/list_changed, notifications/resources/updated, progress, logging) continue to use the transport's existing channels and are unaffected by this proposal.

Lifecycle

• Stream termination. The StreamEventsResult is an empty typed result ({"_meta": {}}). It carries no information — it satisfies JSON-RPC's requirement that every request gets a response. All meaningful content is in the preceding notifications. It is sent whenever the server can write a final frame: on stdio (SHOULD), and on Streamable HTTP only when the server initiates the close.
• Heartbeat. The server MUST send periodic keepalive messages on the push stream so the client can distinguish "nothing to send" from "connection is dead." On Streamable HTTP, this is an SSE comment (: keepalive\n\n). On stdio, this is a notifications/events/heartbeat notification with empty params: {"jsonrpc":"2.0","method":"notifications/events/heartbeat","params":{}}. The server SHOULD send a heartbeat at least every 30 seconds. The client SHOULD treat absence of any data (events or heartbeats) beyond a threshold (e.g., 60 seconds) as connection failure and reconnect with cursors.
• Cancellation. On Streamable HTTP, the client aborts the request stream. On stdio, the client sends notifications/cancelled with the requestId matching the events/stream request's id. In both cases, the server MUST stop delivering events and release any associated resources. On stdio, the server SHOULD then send the StreamEventsResult (the result is empty and harmless; base MCP says servers SHOULD NOT respond to cancelled requests, so a server that omits it is also compliant). On Streamable HTTP, the abort is the terminal signal and no result is sent.
• Updating subscriptions. A client MAY hold multiple concurrent events/stream requests open, each with its own subscription list. To add subscriptions, open an additional stream; to remove them, cancel only the stream that carries them. Nothing prevents a client from instead consolidating onto a single stream by cancelling and re-issuing events/stream with the full updated list — cursor replay covers the transition gap — but this is an optimization, not a requirement. Note that on HTTP/1.1 each stream consumes a TCP connection, so clients that expect many independent subscriptions effectively depend on HTTP/2 multiplexing for the multi-stream approach to scale; SDKs SHOULD coalesce subscriptions onto fewer streams when the transport does not multiplex.
• Reconnection after failure. If the connection drops (HTTP) or the server stops sending (stdio), the client sends a new events/stream with the same subscriptions and their last-known cursors.

Cursor Advancement

Each event notification on the push stream includes a cursor field. This cursor represents the subscription's position after this event. The client tracks the latest cursor per subscription for use during reconnection.

Webhook-Based Delivery

Webhook delivery is for remote MCP servers where maintaining a long-lived connection (push) or frequent polling is impractical. The server POSTs events to a callback URL provided by the client. Webhook subscriptions use soft state with mandatory TTL — the server holds them in memory, but they expire automatically if the client stops refreshing.

The callback URL does not need to be the client itself. A common deployment is a forward proxy that receives webhooks and serves events to clients via poll or push:

Upstream → MCP Server → webhook POST → Forward Proxy ← Client (poll or push)

sequenceDiagram
    autonumber
    participant SDK as Client SDK
    participant Server as MCP Server
    participant Hook as Webhook Endpoint

    SDK->>Server: events/subscribe {id, name, params, delivery: {url}, cursor}
    Server-->>SDK: {secret, refreshBefore}
    loop as events occur
        Server->>Hook: POST {id, event, cursor} + HMAC signature
        Hook-->>Server: 200 OK
        Hook--)SDK: deliver event (implementation-defined,<br/>e.g. forward proxy → poll/push)
    end
    loop before each refreshBefore
        SDK->>Server: events/subscribe (same key — refreshes TTL)
        Server-->>SDK: {refreshBefore'}
    end
    opt explicit teardown (else: stop refreshing → TTL expiry)
        SDK->>Server: events/unsubscribe {id, delivery: {url}}
        Server-->>SDK: (ack)
    end

Subscribing: events/subscribe

Unlike poll and push, webhook delivery requires an explicit subscribe step because the server needs to know where to POST events. events/subscribe is idempotent — calling it again with the same subscription key (see Subscription Identity below) refreshes the TTL and updates mutable fields. This is the mechanism clients use to keep subscriptions alive.

{
  "jsonrpc": "2.0",
  "method": "events/subscribe",
  "id": 2,
  "params": {
    "id": "f47ac10b-58cc-4372-a567-0e02b2c3d479",
    "name": "incident.created",
    "params": { "severity": "P1" },
    "delivery": {
      "mode": "webhook",
      "url": "https://proxy.example.com/hooks/client123"
    },
    "cursor": null
  }
}

// Response
{
  "id": "f47ac10b-58cc-4372-a567-0e02b2c3d479",
  "secret": "whsec_5c8f...",          // present when subscription is (re)created
  "refreshBefore": "2026-02-19T16:30:00Z"
}

Notes:

• events/subscribe is ONLY used for webhook delivery. Poll and push do not need it.
• id is a client-generated, high-entropy identifier for the logical subscription. It MUST contain at least 122 bits of entropy (e.g., a UUIDv4). See Subscription Identity below.
• secret is a shared secret for HMAC-SHA256 signature verification of webhook deliveries. The server generates it when the subscription is created and returns it in the response; it is not returned on subsequent refreshes of an existing subscription. If the server has lost the subscription (restart, TTL expiry) the refresh creates a new one, and a fresh secret appears in the response — the client MUST check for its presence on every refresh and update the verifier accordingly. The client MAY supply delivery.secret in the request to override server generation (e.g., when the secret is provisioned out-of-band in a vault); servers SHOULD accept this but are not required to.
• cursor (request only) tells the server where to begin delivery. null means "start from now." A non-null value requests replay from that position (honoured when the event type is backed by a durable upstream). The cursor is client-owned: the server does not track a delivery watermark, and the response does not include a cursor. The client persists the cursor carried in each delivered payload (see Webhook Event Delivery) and supplies it on every refresh. If the subscription is live, the supplied cursor is at or behind the server's in-flight position and the server treats it as a no-op (delivery continues uninterrupted). If the subscription has lapsed or the server has restarted, the cursor becomes the replay point. This means clients use a single rule — always pass the last-persisted cursor — and the server is idempotent under it.
• refreshBefore is mandatory. It is an ISO 8601 timestamp indicating when the subscription will expire. The client MUST re-call events/subscribe with the same subscription key before this time to keep the subscription alive. The server resets the TTL on each refresh.
• events/subscribe is idempotent within the caller's subscription scope (see Subscription Identity). If a subscription with the same scoped key exists, the server resets the TTL and updates mutable fields in place. If the subscription has expired — or the server has restarted and lost it — the server creates a fresh subscription using the provided cursor.
• The server holds subscription state (id, event name, params, callback URL, secret) in memory with TTL. No durable storage is required — if the server restarts, clients will re-subscribe on their next refresh cycle. For event types backed by a durable upstream, the client's persisted cursor recovers any events that occurred during the gap; for emit-only event types, events during the gap are not recoverable (see Emit-only event types).

Subscription Identity

Webhook subscriptions are keyed by a compound subscription key that determines which events/subscribe calls refer to the same logical subscription. The server uses this key for idempotent upsert on subscribe, for lookup on unsubscribe, and to isolate tenants from one another.

Key composition. The subscription key is:

• On servers that authenticate the caller: (principal, delivery.url, id).
• On servers without caller authentication: (delivery.url, id).

where principal is the server's canonical identifier for the authenticated subject (e.g., OAuth sub, API key ID, service-account name) and delivery.url is the callback URL exactly as supplied by the client. delivery.url is part of the key in both scopes and is therefore immutable for the lifetime of a subscription — to migrate delivery to a new endpoint, the client MUST events/unsubscribe the old subscription and events/subscribe a new one. Servers that support both authenticated and anonymous access MUST select the scoping mode per-request: include principal whenever one is present. A subscription created under one scope is not visible under the other.

id requirements. The id field MUST be a high-entropy value containing at least 122 bits of randomness; a UUIDv4 is RECOMMENDED. The client MUST generate id once per logical subscription and persist it for the subscription's lifetime — a fresh id on each subscribe call creates a new subscription rather than refreshing the existing one. SDKs SHOULD generate and persist id on the client's behalf and SHOULD NOT expose an interface that encourages hand-picked low-entropy values.

Capability semantics (unauthenticated). When the server does not authenticate callers, knowledge of (delivery.url, id) is sufficient to refresh, modify, or delete the subscription. The id therefore functions as a bearer capability and SHOULD be treated as confidential: clients SHOULD NOT log it at default verbosity, embed it in URLs, or expose it to untrusted intermediaries. Servers MAY reject id values that are obviously low-entropy (shorter than 16 bytes, dictionary words, sequential integers) to guard against misconfigured clients.

Mutable vs. immutable fields. name, params, and delivery.url are immutable for a subscription's lifetime: a refresh that supplies different values for any of them addresses a different subscription (the server treats it as a create, not an update). To change what a subscription listens for or where it delivers, the client MUST events/unsubscribe the old one and events/subscribe a new one. This avoids the case where name/params change in place but the upstream listener provisioned by on_subscribe remains bound to the old values.

On an idempotent subscribe against an existing key, the server updates the remaining fields as follows:

Field	Behavior on existing subscription
delivery.secret	Replaced if supplied (client-driven rotation / override). If omitted, the existing secret is unchanged and is not returned.
cursor	The server treats the supplied value as the client's last-persisted position. If the subscription is live and the cursor is at or behind the current in-flight position, this is a no-op. If the subscription has lapsed or the server restarted, delivery (re)starts from this position. The server does not store this value beyond initiating delivery.
TTL	Reset.
active	Set to true. A successful refresh is the client's liveness signal; if delivery had been suspended (active: false in deliveryStatus) due to repeated failures, the server resumes retrying pending events.

Cross-tenant isolation. Because the subscription key always includes delivery.url (and principal when authenticated), two distinct tenants cannot collide on id alone. A malicious caller who learns another tenant's id but not their principal credentials or callback URL cannot refresh, modify, or delete that tenant's subscription, nor redirect its deliveries.

Webhook Event Delivery

The server POSTs events to the callback URL as they occur:

POST https://proxy.example.com/hooks/client123
Content-Type: application/json
X-MCP-Subscription-Id: f47ac10b-58cc-4372-a567-0e02b2c3d479
X-MCP-Signature: sha256=<hex-lowercase HMAC-SHA256(secret, timestamp + "." + body)>
X-MCP-Timestamp: 1739980800

{
  "id": "f47ac10b-58cc-4372-a567-0e02b2c3d479",
  "eventId": "evt_789",
  "name": "incident.created",
  "timestamp": "2026-02-19T16:00:00Z",
  "data": {
    "incidentId": "INC-1234",
    "title": "Database connection pool exhausted",
    "severity": "P1"
  },
  "cursor": "cursor_xyz"
}

Notes:

• X-MCP-Subscription-Id carries the subscription id so the receiver can select the correct secret before parsing the body. X-MCP-Signature contains an HMAC-SHA256 signature (lowercase hex) computed over timestamp + "." + body using the shared secret from subscription. The receiver MUST verify this before processing. The X-MCP-Timestamp header contains the Unix timestamp (seconds) of the request; the receiver SHOULD reject deliveries older than 5 minutes to prevent replay attacks. Each retry attempt regenerates the timestamp and signature.
• eventId in the body enables idempotent processing. The receiver SHOULD deduplicate by this value.
• cursor in the body is the subscription's position after this event. The endpoint MUST make cursor and eventId available to the consuming client by whatever channel it uses to forward events; cursor-based recovery on resubscribe depends on the client receiving and persisting this value.
• Delivery model. The server retries each event independently with exponential backoff on non-2xx responses, and does not track a server-side cursor watermark. This matches the dominant webhook convention (Stripe, GitHub, Shopify, the Standard Webhooks spec). Concurrent deliveries and retries may therefore arrive out of order; the receiver uses eventId for deduplication and timestamp for ordering if needed.
• Acknowledgement semantics. A 2xx response from the webhook endpoint signals that the event has been accepted and the server need not retry it. The endpoint SHOULD NOT return 2xx until the event has been durably persisted or forwarded — an endpoint that ACKs and then loses the event leaves recovery dependent on the client's last-persisted cursor, which may predate the lost event. At-least-once delivery in webhook mode holds between server and endpoint; end-to-end delivery to the agent depends on the endpoint honouring this contract.
• Subscribe/delivery race. Because the server may begin delivering as soon as the subscription is persisted, the first webhook POST can arrive before the events/subscribe response (and thus the secret) reaches the client. The recommended handling for now is retry tolerance: a receiver that gets a delivery for an id it does not yet recognise, or that it cannot yet verify, SHOULD return a retryable status (e.g., 503 or 425 Too Early) rather than dropping it; the server's normal retry/backoff machinery will redeliver once the client is ready. eventId deduplication and cursor replay make this safe. A two-phase activate handshake was considered and may be added later if this proves insufficient in practice.
• After repeated failures (server-defined threshold), the server MAY suspend delivery (deliveryStatus.active: false). A subsequent successful refresh reactivates it (sets active: true) and the server resumes retrying pending events; if the client never refreshes, the subscription expires naturally at TTL.

Webhook Delivery Status

The events/subscribe response MAY include a deliveryStatus object when refreshing an existing subscription. This lets the client detect delivery problems without a separate monitoring channel:

// Healthy subscription refresh
{
  "id": "f47ac10b-58cc-4372-a567-0e02b2c3d479",
  "refreshBefore": "2026-02-19T17:00:00Z",
  "deliveryStatus": {
    "active": true,
    "lastDeliveryAt": "2026-02-19T16:28:00Z",
    "lastError": null
  }
}

// Subscription with delivery failures
{
  "id": "f47ac10b-58cc-4372-a567-0e02b2c3d479",
  "refreshBefore": "2026-02-19T17:00:00Z",
  "deliveryStatus": {
    "active": false,
    "lastDeliveryAt": "2026-02-19T15:45:00Z",
    "lastError": "Webhook endpoint returned 403 Forbidden",
    "failedSince": "2026-02-19T15:50:00Z"
  }
}

deliveryStatus is OPTIONAL — servers MAY omit it entirely. When present, active indicates whether the server is currently delivering events (false means it has suspended retries after repeated failures; the refresh that returned this status has just reactivated it). lastError provides a human-readable description of the most recent failure. The client can use this information to diagnose connectivity or authentication issues with the webhook endpoint.

Webhook Security

SSRF prevention. The server MUST validate callback URLs. Servers SHOULD reject URLs whose resolved IP falls in private, loopback, or link-local address ranges (127.0.0.0/8, 10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16, 169.254.0.0/16, ::1, fc00::/7, fe80::/10) unless explicitly configured to allow them. To prevent DNS rebinding, this validation MUST be performed at delivery time, not only at subscribe time: the server resolves the hostname, checks the resolved IP against the blocklist, and connects directly to that validated IP (sending the original hostname in the Host header / TLS SNI) so the address cannot change between check and connect. Subscribe-time validation alone is insufficient — a rebinding attacker returns a public IP at subscribe time and a private IP at delivery time. Webhook delivery requests MUST NOT follow HTTP redirects, since a redirect can target an internal address that bypasses the blocklist. Servers MAY additionally maintain an allowlist of permitted callback URL patterns.

TLS requirement. Callback URLs SHOULD use https://. Event payloads transit in cleartext over plain http://, exposing them to interception. Servers MAY reject non-TLS callback URLs.

Replay attack prevention. The webhook POST MUST include an X-MCP-Subscription-Id header (the subscription id, so the receiver can select the correct secret without parsing the body) and an X-MCP-Timestamp header containing the Unix timestamp (seconds) at which the request was generated. The HMAC signature MUST cover both the timestamp and the body and is sent as X-MCP-Signature: sha256=<hex-lowercase> where the value is HMAC-SHA256(secret, timestamp + "." + body) encoded as lowercase hexadecimal. Each retry attempt MUST regenerate the timestamp and signature so that retries are not rejected by the receiver's freshness window. The receiver SHOULD reject deliveries where the timestamp is more than 5 minutes old. This prevents captured webhook payloads from being replayed.

Secret generation. The signing secret is generated by the server. This matches the dominant webhook convention (Stripe, Slack, Shopify, the Standard Webhooks spec): the party that signs deliveries owns the key, which guarantees adequate entropy, per-subscription uniqueness, and unilateral rotation. Client-supplied secrets (the GitHub model) are supported as an optional override via delivery.secret for deployments that pre-provision secrets in a vault, but are not the default.

Secret rotation. A client that needs to force rotation calls events/subscribe with the same subscription key and supplies a new delivery.secret; the idempotent upsert replaces it atomically alongside the TTL refresh. Server-initiated rotation happens implicitly whenever the server (re)creates the subscription — the new secret appears in the response and the client adopts it. On unauthenticated servers, the ability to rotate is gated solely by knowledge of (delivery.url, id) — see Subscription Identity for the capability implications.

Authentication to the callback endpoint. HMAC is the only authentication mechanism the protocol defines between the MCP server and the callback endpoint. This matches the dominant pattern for product webhooks (Stripe, GitHub, Slack, Twilio all authenticate deliveries with HMAC alone) and keeps the server's per-subscription credential surface to a single secret.

HMAC is verified by application code after the request has been routed. It does not help if the callback URL sits behind infrastructure that requires transport-level auth before routing — an API gateway that demands Authorization: Bearer, a cloud endpoint requiring a platform-issued identity token, an mTLS-only mesh. For those deployments, point the callback at a forward proxy under the tenant's control: the proxy presents a public endpoint to the MCP server, terminates HMAC, and re-authenticates to downstream services using whatever mechanism the internal network requires. The protocol does not provide a way to pass through bearer tokens, perform an OAuth client-credentials grant, or attach OIDC identity tokens on the tenant's behalf; these may be considered for a future revision if deployment experience warrants it.

Unsubscribing: events/unsubscribe

events/unsubscribe is an eager cleanup mechanism. It is not required for correctness — subscriptions will expire naturally when the client stops refreshing. However, calling it immediately frees server resources and stops webhook deliveries without waiting for TTL expiry.

{
  "jsonrpc": "2.0",
  "method": "events/unsubscribe",
  "id": 3,
  "params": {
    "id": "f47ac10b-58cc-4372-a567-0e02b2c3d479",
    "delivery": { "url": "https://proxy.example.com/hooks/client123" }
  }
}

events/unsubscribe resolves the subscription using the same scoped key as events/subscribe. Both id and delivery.url are required (principal is supplied by the auth layer when present) so the server can form the full key.

events/unsubscribe is ONLY used for webhook delivery. Poll subscriptions are implicit (stop polling to unsubscribe). Push subscriptions are scoped to the events/stream connection (close to unsubscribe).

Listing Active Subscriptions

The client owns all subscription state across all three delivery modes. There is no server-side subscription listing method. The client maintains its own subscription registry and can reconstruct server-side state at any time via idempotent events/subscribe (for webhook mode) or by resuming poll/push with stored cursors.

Enterprise governance tools can inspect the client SDK's subscription registry for audit purposes. For webhook mode, orphaned subscriptions (e.g., from a crashed client) are cleaned up automatically by TTL expiry — no server-side listing is needed.

Cursor Lifecycle

Cursors are opaque strings managed by the server. They represent a position in the event stream.

Initial cursor: Passing cursor: null in a subscription (either in events/poll or events/stream) means "start from now." The server returns a fresh cursor representing the current position. No historical events are replayed.

Cursor expiry: Cursors may become invalid (upstream data compacted, server state reset). The server returns a CursorExpired error for the affected subscription. The client SDK SHOULD handle this by re-subscribing with cursor: null to get a fresh cursor.

For poll mode, this appears in the per-subscription results:

{
  "id": "sub_email",
  "error": {
    "code": -32014,
    "message": "CursorExpired",
    "data": { "reason": "Upstream history compacted" }
  }
}

For push mode, this appears as a notification on the stream:

data: {"jsonrpc":"2.0","method":"notifications/events/error","params":{"id":"sub_email","error":{"code":-32014,"message":"CursorExpired","data":{"reason":"Upstream history compacted"}}}}

For webhook mode, the server POSTs an error envelope to the callback URL (signed like a normal delivery), and additionally surfaces it in deliveryStatus.lastError on the next refresh:

POST https://proxy.example.com/hooks/client123
X-MCP-Subscription-Id: f47ac10b-58cc-4372-a567-0e02b2c3d479
X-MCP-Signature: sha256=<...>
X-MCP-Timestamp: 1739980800

{"id":"f47ac10b-58cc-4372-a567-0e02b2c3d479","error":{"code":-32014,"message":"CursorExpired","data":{"reason":"Upstream history compacted"}}}

On CursorExpired, the client should reconnect the affected subscription with cursor: null.

Server SDK Guidance

The server SDK should make implementing events as simple as implementing tools. The server author writes one function — a "check for changes since cursor" function — and the SDK handles delivery mechanics for all three modes (poll, push, and webhook). The same check function backs every delivery mode; the SDK decides how to deliver based on what the client requested.

By default the server author does not declare which delivery modes an event type supports. The SDK infers them from the server's configuration and transport:

• Poll is available by default — backed by the check function when one is provided, or by the SDK's emit-fed ring buffer otherwise (see Emit-only event types below). Authors MAY opt an event type out of poll (e.g., when buffering is undesirable for a high-volume push-only source).
• Push is available when the transport supports streaming (Streamable HTTP or stdio).
• Webhook is available when the server has webhook_ttl configured.

The SDK computes the delivery array in events/list responses from the above, minus any modes the author has explicitly disabled. The protocol does not require any particular mode to be present, so the SDK MUST also accept an explicit delivery=[...] override per event type for authors who want full control. Servers that call server.emit() for an event type additionally enable real-time push and webhook delivery without the SDK needing to run an internal polling loop.

@server.event(
    name="email.received",
    description="New email arrives",
    input_schema={ ... },
    payload_schema={ ... },
)
async def check_email(context, params, cursor):
    """Called by the SDK for all delivery modes."""
    if cursor is None:
        # Bootstrap: return current position, no events
        current = await gmail.history().get_current_id()
        return EventResult(events=[], cursor=current)

    history = await gmail.history().list(startHistoryId=cursor)
    events = []
    for msg in history.messages:
        if matches_params(params, msg):
            data = {"messageId": msg.id, "from": msg.sender, ...}
            if params.get("redact_pii"):
                data = redact(data)
            events.append(Event(name="email.received", eventId=msg.id, data=data))
    return EventResult(events=events, cursor=history.historyId)

The SDK calls this function in different contexts depending on the delivery mode, but the server author does not need to distinguish between them:

For poll mode: The SDK calls the check function for each subscription in an incoming events/poll request and returns the results. The server is stateless.

For push mode: The SDK supports two patterns:

• Poll-driven push (default). The SDK maintains an internal polling loop per push subscription, calling the check function periodically and writing events to the stream. This is the simplest path for servers wrapping poll-only APIs (e.g., Gmail's history endpoint).
• Direct emit. Servers with true push sources (webhooks, WebSocket listeners, change streams) can emit events directly via server.emit(). The SDK routes emitted events to all active push streams with matching subscriptions.

Push state is scoped to the lifetime of the events/stream connection — when the connection closes, all loops and listeners stop.

For webhook mode: The SDK uses the same two patterns (poll-driven or direct emit), but instead of writing to an SSE stream, it POSTs events to the subscriber's callback URL with HMAC signatures. The SDK holds webhook subscriptions in memory with TTL — no external storage is required. If the server restarts, all webhook subscriptions are lost. Clients will re-subscribe on their next refresh cycle, passing their last-persisted cursor; for event types backed by a durable upstream this resumes without gaps, while emit-only event types lose events that occurred during the outage. This is by design: the mandatory TTL + refresh mechanism eliminates the need for durable subscription storage.

# Webhook mode uses in-memory subscription state with TTL.
# No external storage needed — clients refresh before expiry.
server = MCPServer(
    webhook_ttl=timedelta(minutes=30),  # subscriptions expire after 30 min
)

Direct emit works across all modes. When server.emit() is called, the SDK routes the event to all active subscriptions with matching event name. The SDK supports two emit patterns:

• Broadcast emit. The server emits an event without specifying a subscription. The SDK fans out per active subscription for that event name using two author-supplied hooks:

  • match(ctx, event, params) -> bool decides whether the subscription receives the event. If absent, all subscriptions for that event name receive it.
  • transform(ctx, event, params) -> event shapes the payload for that subscription (e.g., apply redact_pii, expand or thin the payload per a subscriber's expand param). If absent, the event is delivered as emitted.

  ctx carries the subscription's principal and request metadata. Because param semantics are author-defined (filters, globs, transforms — see inputSchema), the SDK cannot evaluate them generically; these hooks let the author own both filtering and shaping. The SDK applies the same hooks when an events/poll request reads from the emit-only ring buffer, so poll subscribers see the same filtering and shaping as push/webhook subscribers. This pattern fits upstream sources that deliver all events regardless of subscription params (e.g., a PagerDuty webhook that fires for every incident — match filters by severity, transform redacts fields per subscriber).

• Targeted emit. The server emits an event to a specific subscription by ID. This is appropriate when the server has set up a per-subscription upstream listener and already knows which subscription the event belongs to.

@server.event(name="incident.created", ...)
class IncidentCreated:
    @staticmethod
    def match(ctx: Context, event: Event, params: dict) -> bool:
        sev = params.get("severity")
        return sev is None or event.data["severity"] == sev

    @staticmethod
    def transform(ctx: Context, event: Event, params: dict) -> Event:
        if params.get("redact_pii"):
            return event.replace(data={**event.data, "reporter": None})
        return event

# Broadcast emit — SDK calls match then transform per active subscription.
async def on_pagerduty_webhook(payload):
    server.emit(Event(
        name="incident.created",
        data={"incidentId": payload["id"], "severity": payload["severity"], ...}
    ))

# Targeted emit — server already knows the subscription.
async def on_slack_message(subscription_id, message):
    server.emit(Event(
        name="slack.message",
        data={"text": message.text, "channel": message.channel, ...}
    ), subscription_id=subscription_id)

Emit-only event types. Many upstream sources are push-only with no cursor-addressable change feed — the server author receives upstream events (via webhook, WebSocket, message bus, etc.) and has no meaningful check() function to write. For these event types, the server author declares the event with emit_only=True and omits the check function entirely:

@server.event(
    name="incident.created",
    description="Fires when a new PagerDuty incident is created",
    emit_only=True,  # no check function; poll served from SDK ring buffer
    input_schema={ ... },
    payload_schema={ ... },
)
class IncidentCreated:
    pass

async def on_pagerduty_webhook(payload):
    server.emit(Event(name="incident.created", data={...}))

SDKs SHOULD provide an in-memory ring buffer that retains a bounded window of emitted events per event type (configurable by time and/or count). events/poll for an emit-only event type reads from this buffer. The cursor is a buffer-local sequence number; it is process-scoped — a server restart invalidates all cursors and clients receive CursorExpired, resetting to "now." Events emitted during server downtime are not recoverable. This matches the upstream's own guarantees for push-only sources and keeps the server author's implementation to a single emit() call per upstream event.

server = MCPServer(
    emit_buffer=EmitBuffer(
        max_age=timedelta(minutes=10),   # retain 10 min of events
        max_events=10_000,               # cap per event type
    ),
)

For servers wrapping upstreams that do offer a durable cursor (Gmail historyId, Kubernetes resourceVersion, Stripe /v1/events, Kafka offset), authors SHOULD prefer a check() function that queries the upstream directly, so the MCP cursor is the upstream cursor and poll survives server restart without gaps.

Subscription lifecycle hooks. For events where the upstream source must be configured per subscription (e.g., "watch this Slack channel"), the server needs to know when subscriptions are added and removed. The SDK provides lifecycle hooks:

@server.on_subscribe("slack.message")
async def on_subscribe(context, params, subscription_id):
    """Set up upstream listener for this subscription's params."""
    await slack.join_channel(params["channel"])

@server.on_unsubscribe("slack.message")
async def on_unsubscribe(context, params, subscription_id):
    """Tear down upstream listener."""
    await slack.leave_channel(params["channel"])

These hooks are called by the SDK across all delivery modes. The server author writes the upstream setup/teardown logic once; the SDK handles the delivery mechanics.

Unsubscribe timing by mode. Push and webhook have explicit end-of-life signals — push fires on_unsubscribe when the stream closes, webhook when events/unsubscribe is called or the TTL lapses. Poll does not: the client simply stops calling events/poll, and the server never sees a goodbye. To prevent poll-provisioned upstream resources from leaking, the SDK treats poll subscriptions as leased. The lease is keyed on (principal, eventName, canonicalHash(params)) — the request id is not used, since it is request-scoped and opaque. on_subscribe fires the first time a given key appears in a poll request; each subsequent poll for that key renews the lease; on_unsubscribe fires when the lease expires without renewal. This lease table is ephemeral SDK state, not protocol state: it is never persisted, and a server restart simply re-fires on_subscribe on the next poll. The lease window is SDK-configurable and SHOULD default to a small multiple of the server's typical nextPollSeconds so that a well-behaved client never lapses between polls:

server = MCPServer(
    poll_subscription_ttl=timedelta(minutes=5),
)

This lease tracking is soft state the SDK maintains on the server author's behalf. It does not survive restart — on restart the next poll is "first sight" again and on_subscribe re-fires, which is the correct behavior since the upstream listener also needs re-establishing. Server authors SHOULD write on_subscribe to be idempotent for this reason.

Client SDK Guidance

Event names and params are not known at code time — they are discovered at runtime via events/list, typically driven by the LLM. The client SDK provides a dynamic subscription API:

client = MCPClient(
    server_url="...",
    webhook=WebhookConfig(                              # optional, global
        url="https://proxy.example.com/hooks/client123",
        # secret is server-generated by default; pass secret=... to override
    ),
)

# Subscribe dynamically (e.g., driven by LLM decision after events/list)
sub = await client.subscribe(
    name="email.received",
    params={"from": "*@anthropic.com", "redact_pii": True},
)

# Receive events
async for event in sub:
    # Called regardless of delivery mode (poll, push, or webhook)
    details = await client.call_tool("get_email", {"id": event.data["messageId"]})
    # ... process the email

# Unsubscribe
await sub.cancel()

The SDK manages delivery mode selection, poll loops, push stream lifecycle, webhook refresh cycles, cursor tracking, deduplication by eventId, cursor expiry recovery, and reconnection — all transparently.

Delivery Mode Selection

The SDK intersects the event type's advertised delivery list with the modes the client is configured to use, and picks in this preference order:

1. Webhook if the client has WebhookConfig set and the server lists it — low latency without tying up a persistent connection.
2. Push if the transport supports streaming (stdio, or HTTP when webhook isn't configured) and the server lists it — low latency, but requires holding a connection open.
3. Poll if the server lists it — no server state, no persistent connection.

If the intersection is empty, client.subscribe() raises NoCompatibleDeliveryMode. The protocol does not guarantee a universal fallback, so client SDKs MUST surface this as an error rather than silently degrading.

The caller can override per subscription:

# Force poll
sub = await client.subscribe("email.received", params={...}, delivery="poll")

# Force push (error if server doesn't support it)
sub = await client.subscribe("email.received", params={...}, delivery="push")

# Force webhook (error if WebhookConfig not set or server doesn't support it)
sub = await client.subscribe("incident.created", params={...}, delivery="webhook")

How Each Mode Works

Poll mode: The SDK runs a background task that calls events/poll at the server-recommended interval, batching all subscriptions to the same server into a single request. Events are yielded to the subscription's async iterator. The caller does not implement or manage the polling loop.

Push mode: The SDK sends an events/stream request and listens for notifications. When subscriptions are added or removed, the SDK either opens an additional stream for the delta or — when the transport doesn't multiplex well (HTTP/1.1) — cancels the current stream and re-issues events/stream with the consolidated list and current cursors. Cursor-based replay covers any transition gap.

Webhook mode: The SDK calls events/subscribe to register the callback URL with the server, which returns the signing secret. It runs a background refresh loop, re-calling events/subscribe before refreshBefore expires; if a refresh response includes a secret, the SDK updates the verifier. The SDK monitors deliveryStatus on each refresh to detect delivery failures. The webhook receiver (which may be a forward proxy) accepts incoming POSTs, verifies HMAC signatures, checks timestamp freshness, and routes events to the appropriate subscription iterator. On sub.cancel(), the SDK calls events/unsubscribe for eager cleanup.

Delivering Events to the LLM

When an event arrives (via any delivery mode), the client SDK surfaces it to the agent runtime. How this works depends on the agent framework — it might inject the event into the conversation context, trigger a new agent turn, or queue it for processing. The MCP spec does not prescribe this; it is an application-level concern.

Security Considerations

Event Payloads Are Untrusted Data

Event payloads MUST be treated with the same caution as tool results. They originate from external systems and may contain:

• Prompt injection attempts (e.g., an email subject line containing "ignore all previous instructions")
• Malformed or malicious data
• PII that should not be forwarded or logged

Clients SHOULD sanitize or sandbox event payloads before presenting them to an LLM. The spec should include guidance similar to the existing tool result security model.

Payload Minimality

Servers SHOULD keep event payloads minimal — enough to identify and triage the event, not the full content. For example, an email event should include sender, subject, and messageId, but not the full email body. The client can use tools to fetch full content when needed.

This reduces PII exposure in event infrastructure (logs, queues, buffers) and limits the injection surface area.

Authorization

Subscribe-time: The server MUST verify that the authenticated user has permission to subscribe to the requested event type with the given params. For example, a Slack server must verify the user has access to the channel specified in the params.

Delivery-time: The server SHOULD periodically re-verify permissions. If the user's access is revoked (e.g., removed from a Slack channel), the server terminates the subscription. The termination signal carries the same nested-error shape across all modes:

Mode	Transport
Push	notifications/events/terminated on the SSE stream
Poll	Error entry in the poll response for that subscription
Webhook	Signed error envelope POSTed to the callback URL, plus deliveryStatus.lastError on next refresh

// Push
data: {"jsonrpc":"2.0","method":"notifications/events/terminated","params":{"id":"sub_slack","error":{"code":-32012,"message":"Unauthorized","data":{"reason":"Access revoked"}}}}

// Webhook (POST body)
{"id":"sub_slack","error":{"code":-32012,"message":"Unauthorized","data":{"reason":"Access revoked"}}}

notifications/events/terminated has the shape {id: string, error: {code: integer, message: string, data?: object}} — identical to notifications/events/error but indicates the subscription has ended, not just a per-event failure. The client SDK SHOULD remove the subscription and notify the application.

Action-time: Event receipt does NOT constitute authorization to act. The agent's response to an event (e.g., calling a tool) goes through normal MCP authorization. The spec should be explicit about this.

Enterprise Governance

The spec does not mandate specific governance mechanisms but is designed to enable them:

• Audit: The client owns all subscription state. Enterprise agent runtimes can inspect the client SDK's subscription registry and log all event-triggered actions.
• Policy: Clients SHOULD support policy evaluation between event receipt and action execution. The policy language is an application concern, not a protocol concern.
• Kill switch: For poll mode, stop polling. For push mode, close the connection (HTTP) or send notifications/cancelled (stdio). For webhook mode, call events/unsubscribe for immediate termination, or simply stop refreshing and the subscription will expire at TTL.
• Rate limiting: nextPollSeconds allows servers to control polling frequency dynamically. Clients SHOULD respect it. Enterprise deployments may impose additional rate limits.

Relationship to Existing Primitives

Resources: notifications/resources/updated remains unchanged. Events are a separate primitive for domain-specific occurrences that don't map to a specific resource. Servers MAY implement resource watching via events, but the two mechanisms are independent.

Tools: Events and tools compose at the application layer. An event arrives, the agent reasons about it, and may call tools in response. The MCP spec does not prescribe event-to-tool wiring.

Prompts: Future work may add event-bound prompts (prompt templates designed to be instantiated when a specific event fires). Not in scope for v1.

Sampling: Events may trigger LLM reasoning. The client decides whether to invoke sampling in response to an event — the server does not control this.

What Is NOT in v1

• Durable subscriptions with replay. No mechanism to replay events from before subscription time. Cursors represent "now" forward.
• Message queue bindings. The protocol does not define bindings for specific message queue systems (SQS, MQTT, Kafka, Pub/Sub, etc.). For high-availability delivery through message queues, deploy an MCP proxy that receives events via poll, push, or webhook and writes to your organization's queue infrastructure. The client reads from the queue using the queue's native client. This keeps the protocol transport-agnostic while enabling any intermediary.
• Rich query language for params. Params are simple key-value. No CEL, JSONPath, or complex predicates. Servers define their own param semantics via inputSchema.
• Cross-server event routing. No mechanism for one server's events to trigger another server's tools. This is an application/orchestration concern.
• Event-bound prompts. Prompt templates that auto-instantiate on events. Deferred to a future version.
• Guaranteed delivery. All three modes provide at-least-once delivery when the cursor is backed by a durable upstream and the client replays from its last known cursor on reconnect/restart. Emit-only event types are at-most-once across server restarts — the in-memory buffer and its cursors do not survive (see Emit-only event types). Exactly-once requires application-level deduplication via eventId.

Open Questions

1. How should the client SDK expose events to agent frameworks? As injected context? As a special message type? As a tool-call-like callback? This is an SDK design question, not a protocol question, but guidance would be useful.

2. Should resource subscriptions and list_changed notifications be re-expressed as events? The existing resources/subscribe → notifications/resources/updated flow and the notifications/{tools,resources,prompts}/list_changed family are effectively a special-cased push-only event channel. Folding them into this primitive (e.g., a reserved mcp.resource.updated event type whose params are {uri}, and mcp.{tools,resources,prompts}.list_changed event types) would give them cursors, poll/webhook delivery, and replay for free, and remove a parallel mechanism from the spec. The cost is a migration path for existing clients and the question of whether the protocol should reserve an mcp.* event-name prefix at all.

3. How should servers publish their egress IP ranges for webhook delivery? Enterprises deploying webhook mode will typically need to allowlist the source IPs that webhook requests originate from. Should the spec define a discovery mechanism (e.g., a well-known endpoint like /.well-known/mcp-egress-ips, or a field in the server's capabilities/metadata)? Or leave it to out-of-band documentation per server? A standard mechanism would simplify enterprise onboarding, but adds burden on server operators who may use dynamic egress (cloud NAT, serverless) where stable IP ranges are hard to guarantee.

4. Should webhook delivery adopt the Standard Webhooks format instead of defining its own? Standard Webhooks specifies a delivery envelope very close to what this doc defines: webhook-id, webhook-timestamp, and webhook-signature headers, HMAC-SHA256(secret, id + "." + timestamp + "." + body) with a whsec_ prefix on secrets, base64 v1,<sig> signature encoding with multi-signature support for rotation, and a versioned scheme that already accommodates asymmetric signing (v1a,). Adopting it would let MCP receivers reuse off-the-shelf verifiers (Svix and others ship libraries in most languages) and would resolve open question 5 below for free. The cost is losing the X-MCP-* header namespace and binding the spec to an external document we don't control. A middle path is to declare MCP webhook delivery a Standard Webhooks profile — same wire format, with MCP-specific body schema and the subscription id carried as webhook-id.

5. Can a single subscription span multiple event names with one cursor? Several upstreams expose one ordered change feed that yields multiple event types — a Kubernetes watch on a namespace produces pod, deployment, and event objects; a Kafka consumer on one topic yields heterogeneous message kinds; a Slack Socket Mode connection delivers messages and reactions. Under the current model, a client wanting both k8s.pod_phase_changed and k8s.oom_killed opens two subscriptions with two cursors, and the server runs two parallel watches against the same apiserver. Options: (a) keep per-event-name subscriptions and let the server SDK internally coalesce upstream connections (no protocol change, SDK complexity); (b) allow a subscription's name to be an array so one cursor covers a set of event types (protocol change, simpler server, client must demux); (c) introduce an event-group concept at registration time. The TypeScript SDK stress-test hit this with both Slack and Kubernetes.

6. Should webhook subscription require an ownership-verification handshake? Before activating delivery, the server would POST a challenge token to delivery.url and require the endpoint to echo it back (cf. Slack's URL verification, SNS SubscriptionConfirmation). This proves the subscriber controls the endpoint, preventing a client from pointing deliveries at a third party. Cost: an extra round-trip and an endpoint-side requirement. The current SSRF defenses (blocklist, no-redirect, delivery-time IP validation) mitigate internal-target abuse but not third-party-target abuse.

7. Should webhook deliveries support multiple signatures for zero-downtime secret rotation? The current design has a single X-MCP-Signature header, and rotation happens via an atomic upsert of delivery.secret. But there's a race: webhooks in flight when the upsert lands were signed with the old secret, and a receiver that has already switched to validating against the new secret will reject them. Stripe-style multi-signature (e.g., X-MCP-Signature: t=<ts>,v1=<sig_old>,v1=<sig_new>) lets the server sign with both secrets during a grace window so the receiver can verify against either. Is the in-flight window small enough to ignore, or should the spec allow delivery.secret to be an array (or require servers to dual-sign for N seconds after rotation)?

Note on Consistency and Ordering

This design intentionally does not provide protocol-level guarantees around event ordering, exactly-once delivery, or transactional consistency. There are no logical clocks, sequence numbers, or cross-subscription ordering constraints.

The rationale is that the protocol connects to many different upstream systems (Gmail, Slack, PagerDuty, etc.), each with their own consistency models. Imposing a unified consistency model at the MCP layer would either be too weak to be useful or too strong to be implementable across diverse backends.

What the protocol does provide:

• Cursors for resumability — the client can pick up where it left off after a disconnect or crash.
• eventId for client-side deduplication — the client can detect and discard duplicates that arise during reconnection. Servers SHOULD set this to the upstream's stable event identifier so dual-path delivery (emit + poll) of the same upstream event collapses to one.
• Per-subscription ordering — for poll and push, events within a single subscription are delivered in the order the server produces them. For webhook, delivery order is best-effort: each event is retried independently and the server does not track a delivery watermark, so concurrent requests and retries can reorder arrivals (this matches Stripe, GitHub, Shopify, and the Standard Webhooks spec). Clients that need ordering use the timestamp field in event payloads. No ordering is guaranteed across subscriptions in any mode.

Servers that need stronger guarantees can implement them. For example, a server wrapping a Kafka topic could expose partition offsets as cursors and provide exactly-once semantics within a partition. A database change-data-capture server could use LSNs as cursors and guarantee causal ordering. The protocol's cursor mechanism is flexible enough to support these — the cursor is opaque, so it can encode whatever the server needs (sequence numbers, timestamps, composite positions).

The client and agent layer should be designed to tolerate out-of-order and duplicate events. For most AI agent use cases — reacting to new emails, triaging alerts, responding to messages — eventual consistency with deduplication is sufficient. Agents that need stronger guarantees should use tools to read authoritative state rather than relying solely on event payloads.

Note on Flow Control

This design does not include a protocol-level flow control mechanism (e.g., credit-based or token-based backpressure). This is intentional for v1.

In practice, the dominant bottleneck in an MCP event pipeline is LLM inference — processing a single event may take seconds to minutes of model time. Compared to this, event delivery rates from typical upstream sources (email, Slack, PagerDuty) are negligible. The system is inherently consumer-bound, not producer-bound.

What exists today is sufficient:

• Transport-level backpressure. TCP flow control naturally throttles the server when the client stops reading from the connection. On stdio, OS pipe buffers provide the same effect.
• Server-controlled push frequency. For push mode, the server controls how often it checks upstream sources and delivers events. For poll mode, nextPollSeconds lets the server throttle the client explicitly.
• Client-side prioritization. When multiple events arrive while the LLM is busy, the client SDK can prioritize by event type, severity, or recency. This is an SDK concern, not a protocol concern.

Head-of-line blocking — where a slow-to-process event delays delivery of subsequent events — is likewise handled at the client SDK layer. The SDK can maintain per-subscription queues and let the agent framework process events concurrently or by priority, rather than strictly in arrival order.

One known gap is push-mode reconnect replay: a client reconnecting with a stale cursor may receive a large backlog burst on the stream, with no protocol-level bound equivalent to poll's maxEvents/hasMore. For v1 this is left to TCP backpressure and server-side pacing. If future use cases involve high-throughput event streams where producer-side backpressure or bounded replay becomes necessary, a credit-based flow control extension can be added without breaking the existing protocol.

Appendix: End-to-End Example — GitHub MCP Server (Webhook Delivery)

This walks one concrete topology from upstream source to agent reaction, using webhook delivery via a forward proxy. The MCP server wraps the GitHub API: it receives PR activity from GitHub via GitHub's native repository webhooks (the upstream channel — outside this protocol), and delivers it to a forward proxy via MCP webhook (events/subscribe). The proxy buffers events and serves them to the client over poll or push. When an event arrives, the client SDK wakes the LLM agent.

sequenceDiagram
    autonumber
    participant LLM as LLM Agent
    participant SDK as Client SDK
    participant Proxy as Forward Proxy<br/>(webhook endpoint)
    participant Server as GitHub MCP Server
    participant Redis as Redis<br/>(sub store, TTL)
    participant GH as GitHub.com<br/>(upstream)

    rect rgb(245,245,250)
        Note over SDK,Redis: Subscribe (webhook delivery)
        SDK->>Server: events/subscribe<br/>{id, name: "pull_request.opened", params: {repo: "acme/webapp"},<br/>delivery: {mode: "webhook", url: "https://proxy/.../hooks"}}
        Server->>Redis: SETEX sub:{principal}:{urlHash}:{id} {ttl}<br/>{name, params, url, secret}
        Server->>GH: ensure repo webhook registered<br/>(server-internal, idempotent)
        Server-->>SDK: {secret, cursor, refreshBefore}
    end

    rect rgb(245,250,245)
        Note over GH,LLM: Event fires
        GH->>Server: POST /gh-webhook<br/>{action: "opened", pull_request: {...}}
        Server->>Redis: SCAN sub:* → match repo
        Redis-->>Server: [{id, url, secret}]
        Server->>Proxy: POST {id, event: {number: 42, title, author, url}, cursor'}<br/>+ HMAC signature
        Proxy-->>Server: 200 OK
        Proxy--)SDK: deliver event<br/>(implementation-defined, e.g. poll/push)
        SDK->>LLM: invoke agent with event payload
    end

    rect rgb(250,245,245)
        Note over SDK,Redis: Keepalive
        SDK->>Server: events/subscribe (same key — refresh TTL)
        Server->>Redis: EXPIRE sub:{principal}:{urlHash}:{id} {ttl}
        Server-->>SDK: {refreshBefore'}
    end
    Note over Redis: no refresh → key expires,<br/>subscription gone

The Redis lane is illustrative, not normative. The body of this doc says webhook subscriptions are "in memory with TTL"; that is sufficient for a single-process server. A horizontally scaled server (multiple replicas behind a load balancer) needs shared state so any replica can match an incoming upstream event to subscriptions created on another replica — Redis with key TTL is the obvious fit, but any shared store with expiry works. The protocol does not require this state to be durable: if it is lost, clients re-create it on their next refresh and the client's persisted cursor covers the gap (for event types backed by a durable upstream).

Note there are two webhook hops here, and they are unrelated: GitHub → MCP Server (step 4) is GitHub's native webhook product, configured server-side and outside this spec; MCP Server → Forward Proxy (step 5) is the MCP events/subscribe webhook defined above.

Substituting other delivery modes. Only the Subscribe and Event fires boxes change; the upstream leg (GitHub → MCP Server) is identical in all three modes.

• Push — step 1 becomes events/stream; drop the Forward Proxy lane and the Keepalive box; steps 5–7 collapse into a single notifications/events/event from server to SDK (see Push-Based Delivery).
• Poll — drop step 1 and the Forward Proxy lane; the SDK loops events/poll with the subscription inline, and step 5–7 become the poll response (see Poll-Based Delivery).