Rivery MCP server
Yes. Rivery publishes an official MCP server at https://mcp.rivery.io/mcp. It's hosted. It offers 32 tools. We last checked it on 22 Nov 2026.
Official: answers MCP requests at mcp.rivery.io, on Rivery's own domain rivery.io. Checked 22 Nov 2026. Maintained by Stuart Brameld.
Key facts
As of 22 Nov 2026, Rivery's official MCP server scores 55/100 (provisional) on MartechDB's Agent Readiness Index (v1.1). It answered 55 of our 55 checks in the last 90 days and lists 32 tools.
How the Agent Readiness Index works · Wrong? Suggest a correction
Rivery MCP server review
55/100
Rivery's MCP server scores 55/100 on MartechDB's Agent Readiness Index: usable by AI agents, with gaps.
The good
Hosted by Rivery: connect by URL, with nothing to install
Answered 100% of our checks in the last 90 days
The bad
Nothing we've found.
Not checked yet: how you sign in; a docs page listing its tools; what agent access costs. Anything not checked scores nothing.
Official MCP server
Rivery MCP server
- Address
- https://mcp.rivery.io/mcp
- Runs
- Hosted
- Sign-in
- Official MCP Registry
- Not listed
- How we know it's official
-
- Answers MCP requests at mcp.rivery.io, on Rivery's own domain rivery.io.
- Tools
- 32 tools · listed 21 Nov 2026
- First seen
- 1 Oct 2026
- Last checked
-
Reachable ·
16 Nov 2026: reachable17 Nov 2026: reachable18 Nov 2026: reachable19 Nov 2026: reachable20 Nov 2026: reachable21 Nov 2026: reachable22 Nov 2026: reachable
- Uptime
- 100% · 55 of 55 checks reachable since 1 Oct 2026
Rivery MCP server tools (32)
As the server lists them. Names and descriptions are the vendor's own.
-
get_contextReturn the configured Boomi Data Integration account and environment defaults for this server. Call this first to discover which account_id and environment_id are pre-configured. If they are set, you can omit them from other tool calls. AUTHENTICATION: The token is configured once as the `X-Rivery-Token` header in the user's MCP client config (a Boomi JWT or a Rivery token) — you never pass a token to any tool, and it never appears in the conversation. With it set, tools just work and account_id/environment_id are populated automatically. If a tool fails with "Missing required auth", tell the user to add the `X-Rivery-Token` header to their MCP config; you can call get_rivery_token to verify it resolves. CRITICAL TERMINOLOGY RULES — NEVER break these with the user: - NEVER say "Rivery" or "river" to the user. The product is "Boomi Data Integration". - NEVER say "river" for a pipeline. Always say "data flow". - If the user says "river" or "Rivery", do NOT repeat those words back. Silen…
-
get_rivery_tokenConfirm authentication and report the active account/environment. The token is NOT passed here. It is read from the `X-Rivery-Token` header that the user sets once in their MCP client config. It may be a Boomi JWT or a Rivery token — the type is detected automatically and a Boomi JWT is swapped server-side; the raw token never leaves the server or enters this response. Call this to verify auth is working and to discover which account_id / environment_id the configured token resolves to. If it reports authenticated: false, tell the user to add the `X-Rivery-Token` header to their MCP config. environment_id (optional, Boomi swap only) scopes the swapped token to that environment; otherwise the account default is used.
-
list_accountsList Boomi Data Integration accounts accessible with this token. ALWAYS call with no arguments first — returns up to 5 accounts. Show the list to the user and ask: "Is your account here? If not, tell me your account name and I'll search for it." If the user provides a name not in the list, call again with search=<name>. Once the user confirms their account by name, call setup_account. Never auto-select — always confirm the account name with the user.
-
setup_accountPersist an account selection for the current session and return the config snippet. Call this only after the user has explicitly confirmed their account name. Show the returned config_snippet to the user and ask them to add it to their MCP env block — once set, account discovery never runs again.
-
list_data_flowsList data flows by name or status. Do NOT call this speculatively. Before calling, ask the user what they want to do. You can: - Run a data flow → ask for the name, then call list_data_flows(name=<name>) - Check run status → ask which flow and use get_run / list_runs - Create a new flow → use create_data_flow - Update a flow → ask for the name, then get it and patch it Always filter by name when the user mentions a specific flow. Only list without a filter if the user explicitly asks to see all flows. Returns items plus pagination metadata (total_items, page, has_next).
-
get_data_flowGet full details of a data flow by its ID. Note: only works for data flows created via the API (is_api_v2=true). Data flows created in the Boomi Data Integration UI will return a 400 error. Use list_data_flows to find data_flow_id values.
-
create_data_flowCreate a new source-to-target data flow.
-
update_data_flowUpdate an existing data flow. body must match the UpdateRiver API schema. To add source tables, use set_data_flow_source_tables instead — it handles the correct schema grouping automatically.
-
set_data_flow_source_tablesSet the source tables on a data flow by name. Looks up the correct schema grouping from the connection and builds the data flow schemas structure automatically. Use this instead of update_data_flow when you want to select which tables to sync. Note: this performs a GET-then-PUT with no locking. Concurrent calls on the same data_flow_id will overwrite each other (last write wins). Avoid calling this on the same data flow from multiple agents simultaneously.
-
delete_data_flowDelete a data flow. The upstream DELETE returns an empty body on success; this returns {"deleted": true, "data_flow_id": ...}. Deleting an ACTIVE flow returns a 400 (surfaced verbatim) — call disable_data_flow first, then delete. This never force-deletes.
-
run_data_flowTrigger an immediate run of a data flow. By default returns instantly with the run_id (do NOT poll). Set wait=True to block until the run reaches a terminal state (success/failed/cancelled) and return the final run object — useful for tests/automation so you don't sleep blindly. Before calling this, make sure ALL source tables are configured in a single set_data_flow_source_tables call — one data flow syncs multiple tables in one run. Common 400 causes: - Data flow not yet activated: call activate_data_flow(wait=True) first. - No source tables configured: call set_data_flow_source_tables first.
-
activate_data_flowActivate (schedule) a data flow so it runs on its configured schedule. Set wait=True to block until activation completes or fails — eliminates the need to manually call get_operation in a loop. Returns the final operation status when wait=True.
-
disable_data_flowDisable (pause) a data flow's schedule without deleting it. Returns {"disabled": true, "data_flow_id": ...} (the upstream may return an empty body).
-
patch_data_flowPartially update a data flow — only supply the fields you want to change. Fetches the current data flow, deep-merges patch into it, then PUTs it back. Much simpler than update_data_flow when you only want to change one or two fields. IMPORTANT — lists are REPLACED, not merged. Do NOT use this to add or change source tables (properties.schemas). Use set_data_flow_source_tables instead, which handles table selection correctly and keeps all tables in one data flow. Examples -------- Change target schema only: patch={"properties": {"target": {"schema_name": "PUBLIC"}}} Rename the data flow: patch={"name": "New Name"} Note: GET-then-PUT with no locking — concurrent calls on the same data flow will overwrite each other (last write wins).
-
list_connectionsList all connections in an environment (sweeps every page in one call). Returns {items, total_items, page, has_next}. Each item includes connection_name, connection_type, connection_type_id, and cross_id; use cross_id as connection_id when creating data flows. If the upstream 500s on a specific page, that page is skipped and its number is reported under "incomplete_pages" rather than failing the whole listing.
-
get_connectionGet a single connection by its cross_id.
-
create_connectionCreate a new connection. body must match the CreateConnection API schema. NOT ALL CONNECTION TYPES CAN BE CREATED HERE. Connections that authenticate via OAuth (e.g. Google services, Salesforce, HubSpot, Facebook/Meta, LinkedIn, and similar — anything that needs a browser sign-in / "Connect with..." flow) CANNOT be created through this tool, because OAuth requires an interactive browser consent step. For those, tell the user: "This connection uses OAuth — please create it in the Boomi Data Integration console, then I can use it here." This tool works for connections that authenticate with credentials you can supply directly (host/port/user/password/keys), e.g. databases like mysql, postgresql, mssql, snowflake, bigquery, redshift, mongodb.
-
update_connectionUpdate an existing connection.
-
delete_connectionDelete a connection. The upstream DELETE returns an empty body on success; this returns {"deleted": true, "connection_id": ...}.
-
get_connection_schemasList available schemas/databases for a connection. If the schemas endpoint reports none (metadata not loaded yet), falls back to deriving distinct schema names from the connection's cached table metadata and flags the result with derived_from_tables=true. If still empty, returns a note suggesting reload_connection_metadata.
-
get_connection_tablesList available tables for a connection, optionally filtered by schema. IMPORTANT: this returns CACHED metadata, not a live probe. A successful response does NOT mean the source is reachable — a connection whose DB is down can still return stale tables. Use last_metadata_refresh (newest item updated_at) to judge staleness; the only real liveness signal is actually running the data flow. Returns the first page only. For connections with many tables (e.g. Salesforce), use set_data_flow_source_tables directly — it resolves table names via targeted lookups.
-
reload_connection_metadataTrigger a background metadata refresh (schemas and tables) for a connection. Submits a get_db_metadata pull request to the v1 API (POST /accounts/{a}/environments/{e}/pull_requests). The source type (e.g. hubspot_v3) is normalized to the external datasource name the API expects (hubspot); the API maps it back to the internal id itself. An unrecognized type raises a clear error (not a 422). source_name: the connection's type (e.g. "hubspot_v3"/"hubspot", from list_connections). Pass it to skip the connection lookup; if omitted, the type is read from the connection. Use when get_connection_tables / get_connection_schemas come back empty. Runs asynchronously — tell the user to retry get_connection_tables in a minute or two; do NOT poll in a loop. Returns {"reload_started": true, connection_id, datasource_id, operation_id?, run_id?}; the operation_id can be checked with get_operation.
-
list_source_examplesBrowse available source-to-target integration examples. Two usage modes: **Hierarchical drill-down** (step by step): - No args → list all integration_types - integration_type only → list sources for that type - integration_type + source → list reports - integration_type + source + report → list targets **Cross-cutting filter** (source or target alone, without integration_type): - source="adobe_analytics" → all complete paths that use that source - target="snowflake" → all complete paths that target snowflake - source=X, target=Y → paths matching both Use the results to build arguments for get_source_example.
-
get_source_exampleLoad a source-to-target example configuration. example_type: "post" (request body for creating/updating) or "get" (full API response shape). Default is "post" — use it as the body template for create_data_flow or update_data_flow. Raises FileNotFoundError if the combination does not exist in the library.
-
list_environmentsList all environments for an account.
-
list_variablesList all variables in an environment.
-
upsert_variablesCreate or update variables in an environment. Each item in variables should be: {"name": "VAR_NAME", "value": "some_value"}. Returns {"updated": true, "count": N} (the upstream may return an empty body).
-
get_operationGet an activation/validation operation by its operation_id (24-char hex). This is for activate_river / validate operations only. For data run status, use get_run instead.
-
list_data_flow_operationsList activation/validation operations for a data flow. This shows activate_data_flow and validate history only. For data run history, use list_runs instead.
-
get_runGet the status and result of a data flow run by its run_id. Use the run_group_id returned by run_data_flow as the run_id here. Status is returned as a plain string: pending, running, success, failed. Check error_description for failure details.
-
get_run_logsDownload the logs for a specific data flow run as CSV text. Returns raw CSV content with all log lines for the run. Logs are only available for runs within the last 2 weeks. If the response is empty, the data source may not support logs. Use get_run first to confirm the run exists, then call this to fetch its logs.
-
list_runsList data runs for a data flow within a time range. start_time and end_time must be ISO 8601 format, e.g. "2026-05-25T00:00:00". Each item includes run_id, status, start_date_utc, end_date_utc, error_description.
Connect your AI to Rivery
Add the official Rivery MCP server to Claude, ChatGPT or your editor.
Connect Rivery MCP to Claude
Claude.ai and Claude Desktop
Open Customize, then Connectors.
Click +, then Add custom connector.
Name it Rivery, paste the address below and click Add.
Connect Rivery MCP to ChatGPT
Open Settings, then Apps, then Advanced settings.
Turn on Developer mode, then click Create.
Name it Rivery and paste the address below.
ChatGPT usually only connects to servers with self-serve sign-in or none.
Connect Rivery MCP to Cursor
Opens Cursor if it's installed. If nothing happens, use the JSON below.
Connect Rivery MCP to VS Code
Opens VS Code if it's installed. Otherwise add the JSON settings from the Cursor tab to your .vscode/mcp.json.
Rivery MCP server questions
Does Rivery have an official MCP server?
Yes. Rivery publishes its MCP server itself. Its address is on Rivery's own domain (mcp.rivery.io). MartechDB only lists servers a vendor publishes, never community copies.
What is the Rivery MCP server URL?
The address is https://mcp.rivery.io/mcp.
How do I connect Rivery MCP to Claude, ChatGPT or Cursor?
In Claude, add https://mcp.rivery.io/mcp as a custom connector (Customize, Connectors, Add custom connector). In Cursor, use the Add to Cursor link on MartechDB's page, or paste the JSON settings into your mcp.json. In Claude Code, run: claude mcp add --transport http rivery https://mcp.rivery.io/mcp
What can the Rivery MCP server do?
It offers 32 tools, including get_context, get_rivery_token, list_accounts, setup_account and list_data_flows (as listed on 21 Nov 2026).
Is the Rivery MCP server in the official MCP Registry?
No. Rivery hasn't published it to the official MCP Registry.
Is the Rivery MCP server working?
Yes. It answered our most recent check. We last checked it on 22 Nov 2026.
What are the alternatives to Rivery's MCP server?
In Data integration, the most agent-friendly alternatives with an official MCP server are Needle (97/100), Supermetrics (97/100), Activepieces (97/100), IFTTT (92/100) and import.io (92/100). Rivery scores 55/100 on the same index.
Alternatives to Rivery's MCP server
Vendors in Data integration with an official MCP server, the most agent-friendly first. Rivery scores 55/100.
Needle
needle.app
Supermetrics
supermetrics.com
Activepieces
activepieces.com
IFTTT
ifttt.com
import.io
import.io
Use with AI
Connect your AI to MartechDB
Then ask it about Rivery, or any of the 1,850 marketing tools with an official MCP server. Free, no sign-in.
Claude.ai and Claude Desktop
Open Customize, then Connectors.
Click +, then Add custom connector.
Name it MartechDB, paste the address below and click Add. There's no sign-in.
The free plan allows one custom connector. On Team and Enterprise plans, an owner adds the connector first.
Claude Code
Run this in your terminal:
ChatGPT
Open Settings, then Apps, then Advanced settings.
Turn on Developer mode, then click Create.
Name it MartechDB, paste the address below and choose No authentication.
Developer mode is on Plus and Pro plans, and on Business, Enterprise and Edu if an admin allows it.
API
The same data as JSON, with no sign-in. See the OpenAPI description for every endpoint and field.
CLI
Search the directory and print the command to connect a vendor's server, from your terminal. Needs Node 18 or later. View on npm
Show that your MCP server is verified
If you work at Rivery, add this badge to your MCP docs or README. It's free, it's the same for every vendor, and it updates from our daily checks: it shows uptime while the server responds, and says so if it stops.
Something wrong or out of date? Suggest a correction or email us, and we'll check it against the evidence. Every correction we make is listed on Corrections.