M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
C

Confluent MCP Server

Official
An open-source MCP server that enables AI assistants to interact with Confluent Cloud, Confluent Platform, and Apache Kafka through natural language.
How to install ↓ Compare GitHub source repository ↗
Category
Database #33 of 58
Stars
★ 167 Very popular
Transport
stdio (local process) · SSE · Streamable HTTP
Runtime
Node.js 22.19.0+
Credentials
Optional API key
License
MIT
Last commit
Tools
73
54FMRS · D

Confluent's official open-source MCP server, feature-rich covering Kafka, Flink, Schema Registry, etc., but configuration is complex and community support is limited.

Strongest · Documentation 14/20 Weakest · Reliability 6/20

Reliability
6/20
Security and permissions
9/20
Maintenance
12/20
Documentation
14/20
Setup experience
13/20
Why each score
Reliability 6/20
Only the README and repository metadata are available; there is no verifiable evidence that the server starts cleanly or completes the MCP init and tool-list handshake. The README claims unit/integration tests and a Docker Compose smoke test, but no CI workflow or committed test files were provided. The happy path appears plausible (tool list and CLI output are internally consistent), but the lack of reproducible execution evidence, thin error-handling details, and uncertainty require a conservative score.
Security and permissions 9/20
Credentials are handled via YAML, environment variables, static API keys, and OAuth PKCE; sample configs use placeholders rather than real secrets. Telemetry via Sentry is disclosed with an opt-out (DO_NOT_TRACK), and --disable-auth carries an explicit warning. DNS rebinding protection via --allowed-hosts is also present. However, destructive tools such as delete-topics, delete-schema, delete-connector, and delete-flink-statements exist without any documented confirmation mechanism. Least-privilege scoping and data-flow disclosure are not fully demonstrated, so the score is moderate-high rather than full.
Maintenance 12/20
The repository is not archived, carries a clear MIT license, and shows npm distribution signals (version badge), indicating ongoing maintenance. However, there are 41 open issues, the README explicitly states community-supported best-effort support with no SLA, and no security-response channel, dependency update policy, or changelog evidence is visible. This is active maintenance with governance and versioning gaps, so a score of 12 reflects sustained but imperfect upkeep.
Documentation 14/20
The README is layered and covers quick start, full configuration, OAuth, CLI usage, examples for multiple mainstream MCP clients, telemetry disclosure, and troubleshooting, with links to supplemental documents such as CONFIGURATION.md. Sample configurations use placeholders and the tool tables are detailed. However, there are no explicit cost or rate-limit sections, and all documentation comes from untrusted evidence that cannot be independently verified, so full marks are not given.
Setup experience 13/20
The setup path is clearly described: Node.js 22.19+ prerequisite, npx/pnpm execution, --init-config to generate a config, editing the YAML, then running with --config. Client-specific guides cover Claude Code/Desktop, Cursor, VS Code, Goose, Gemini CLI, and others. Platform compatibility and prerequisite notes are clear. Still, users must manually populate credentials and service endpoints, setup is moderately complex, and there is no verifiable execution evidence in this static review; per the calibration rule setup cannot exceed 15, so 13 is appropriate.

Static review · not runListed 2026-08-07

Read the FMRS scoring method →

Fit and risk

What it can accessUses the networkChanges third-party account data

Best for

  • AI assistants interacting with Confluent Cloud or Kafka
  • Development environments using Confluent Platform or local Kafka
  • Teams needing a unified toolset for Kafka, Flink, Schema Registry

Not for

  • Kafka users outside the Confluent ecosystem (only subset of tools available)
  • Users wanting a fully managed MCP server (consider Confluent's managed service)
  • Users requiring dedicated support (community support is best-effort only)

Required permissions

  • Access Kafka topics, consumer groups, and configurations
  • Execute Flink SQL statements and read results
  • Manage schemas in Schema Registry
  • Manage connectors and Tableflow topics
  • Read Confluent Cloud metrics and billing data

Risks and side effects

  • Destructive operations (e.g., delete topics, delete schemas) must be authorized carefully
  • OAuth credentials might be exposed if config files are not protected
  • HTTP/SSE transports without API key may be vulnerable to unauthorized access
  • Tableflow tools require cloud IAM permissions; misconfiguration may cause authorization errors

Setup

Before you start

Runtime：Node.js 22.19.0+

KAFKA_API_KEY optionalsecret Confluent Cloud Kafka API key, created in the Confluent Cloud console; used for Kafka auth on direct connections via ${KAFKA_API_KEY} interpolation in sample configs.
KAFKA_API_SECRET optionalsecret Secret paired with KAFKA_API_KEY, also generated in the Confluent Cloud console.
SCHEMA_REGISTRY_API_KEY optionalsecret Schema Registry API key used for HTTP Basic Auth against Schema Registry.
SCHEMA_REGISTRY_API_SECRET optionalsecret Secret paired with SCHEMA_REGISTRY_API_KEY.
MCP_API_KEY optionalsecret API key for HTTP/SSE transport auth; generate with --generate-key and put it under server.auth.api_key in config.yaml.
CP_KAFKA_PASSWORD optionalsecret Kafka password for running the Confluent Platform integration tests (default mcp-secret in the docker-compose test stack).
Other optional settings (3)
NODE_EXTRA_CA_CERTS optional Path to an internal CA certificate bundle; set it before starting the server if you see TLS handshake failures.
DO_NOT_TRACK optional Set to true to opt out of telemetry collection (Sentry error reporting and usage data).
CP_KAFKA_USERNAME optional Kafka username for running the Confluent Platform integration tests (default mcp in the docker-compose test stack).
  1. Ensure Node.js 22.19.0 or later.
  2. Generate a config file with npx @confluentinc/mcp-confluent --init-config.
  3. Edit the config.yaml file.
  4. Start the server with npx @confluentinc/mcp-confluent --config ./config.yaml.
  5. Configure your MCP client to connect.

Check that it works

Run npx @confluentinc/mcp-confluent --list-tools to see enabled tools; after connecting your client, the always-available tools like list-configured-connections should appear, and calling it confirms your config connections are loaded.

Troubleshooting

  1. Node.js version too low: upgrade to 22.19.0+
  2. Tools not appearing: run `--list-tools` to check enabled tools, or call `explain-disabled-tools`
  3. Authentication errors on HTTP/SSE: generate an API key with `--generate-key` and set it in `server.auth.api_key`
  4. Port conflicts: change `server.http.port` in config.yaml
  5. Tableflow authorization errors: follow the setup for IAM roles and policies

Things to try

Once connected, you can ask your AI assistant things like:

  • List all topics in my Kafka cluster
  • Show me the consumer group lag across all groups
  • Search the Confluent docs for Tableflow
  • Create a topic named orders with 3 partitions and replication factor 3

Tools 73

search-product-docs read-only
Search Confluent product documentation.
get-product-doc-page read-only
Fetch the full markdown content of a Confluent product documentation page.
explain-disabled-tools read-only
Explain why tools are missing or unavailable.
list-configured-connections read-only
List every configured connection and the tools available on each.
config-help read-only
Suggest configuration to enable a specific tool.
describe-configured-connection read-only
Describe one connection's config and tool availability.
list-topics read-only
List all topics in the Kafka cluster.
create-topics writes
Create one or more Kafka topics with optional partition count and replication factor.
Show 65 more tools
delete-topics destructive
Delete topics with the given names.
produce-message writes
Produce records to a Kafka topic, supporting Confluent Schema Registry serialization.
consume-messages read-only
Consume messages from Kafka topics, optionally restricted to partition, offset, etc.
list-consumer-groups read-only
List consumer groups on a Kafka cluster.
describe-consumer-group read-only
Describe a single consumer group, returning state, type, protocol, etc.
get-consumer-group-lag read-only
Compute live offset lag for a Kafka consumer group.
alter-topic-config writes
Alter topic configuration in Confluent Cloud.
get-topic-config read-only
Retrieve configuration details for a Kafka topic.
get-partition-offsets read-only
Return per-partition low/high watermarks and message counts for a topic.
create-flink-statement writes
Create a Flink SQL statement.
list-flink-statements read-only
List Flink SQL statements.
get-flink-statement-results read-only
Fetch result rows produced by a Flink SQL statement.
delete-flink-statements destructive
Delete Flink SQL statements.
get-flink-statement-exceptions read-only
Retrieve the most recent exceptions for a Flink SQL statement.
list-compute-pools read-only
List Flink compute pools.
list-flink-catalogs read-only
List catalogs available in the Flink environment.
list-flink-databases read-only
List all databases (schemas) in a Flink catalog.
list-flink-tables read-only
List all tables in a Flink database.
describe-flink-table read-only
Get full schema details for a Flink table.
get-flink-table-info read-only
Get table metadata via INFORMATION_SCHEMA.TABLES.
check-flink-statement-health read-only
Perform an aggregate health check for a Flink SQL statement.
detect-flink-statement-issues read-only
Detect issues for a Flink SQL statement.
get-flink-statement-profile read-only
Get Query Profiler data for a Flink SQL statement.
list-connectors read-only
List active connectors.
get-connector-config read-only
Retrieve the full configuration map for a connector.
get-connector-offsets read-only
Retrieve current offsets for a connector's tasks.
get-connector-status read-only
Get the current state of a connector and its tasks.
get-connector-tasks read-only
List the tasks of a connector along with their configurations.
get-connector-error-summary read-only
Summarize a connector's current errors.
get-connector-error-recommendations read-only
Get suggested remediation steps for a failed connector.
get-connector-logs read-only
Retrieve recent log entries for a Confluent Cloud connector.
create-connector writes
Create a new connector.
delete-connector destructive
Delete an existing connector.
pause-connector writes
Pause a running connector and its tasks.
resume-connector writes
Resume a paused connector and its tasks.
restart-connector writes
Restart a connector and its tasks.
update-connector-config writes
Update the configuration of an existing connector.
list-schemas read-only
List all schemas in the Schema Registry.
create-schema writes
Register a new schema or a new version under a subject.
delete-schema destructive
Delete a schema subject or a specific version from the Schema Registry.
search-topics-by-tag read-only
List all topics in the Kafka cluster with the specified tag.
search-topics-by-name read-only
List all topics in the Kafka cluster matching the specified name.
create-topic-tags writes
Create new tag definitions in Confluent Cloud.
delete-tag destructive
Delete a tag definition from Confluent Cloud.
remove-tag-from-entity writes
Remove tag from an entity in Confluent Cloud.
add-tags-to-topic writes
Assign existing tags to Kafka topics.
list-tags read-only
Retrieve all tags with definitions from Confluent Cloud Schema Registry.
list-organizations read-only
List Confluent Cloud organizations the current credentials can see.
list-environments read-only
Get all environments in Confluent Cloud.
read-environment read-only
Get details of a specific environment by ID.
list-clusters read-only
Get all clusters in the Confluent Cloud environment.
create-tableflow-topic writes
Create a Tableflow topic.
list-tableflow-topics read-only
List Tableflow topics.
read-tableflow-topic read-only
Read a Tableflow topic.
update-tableflow-topic writes
Update a Tableflow topic.
delete-tableflow-topic destructive
Delete a Tableflow topic.
list-tableflow-regions read-only
List Tableflow regions.
create-tableflow-catalog-integration writes
Create a catalog integration.
list-tableflow-catalog-integrations read-only
List catalog integrations.
read-tableflow-catalog-integration read-only
Read a catalog integration.
update-tableflow-catalog-integration writes
Update a catalog integration.
delete-tableflow-catalog-integration destructive
Delete a catalog integration.
list-available-metrics read-only
List available Confluent Cloud metrics.
query-metrics read-only
Query Confluent Cloud metrics.
list-billing-costs read-only
Query billing and cost data.

Use cases

Manage Kafka topics, consumer groups, and messages
Run Flink SQL queries and fetch results
Manage Schema Registry, connectors, and Tableflow
Query Confluent Cloud metrics and billing information

Supported clients

Claude Desktop
Claude Code
Cursor
VS Code
Goose
Gemini CLI

Listed from the project's documentation, not tested by this site.

Overview

Confluent MCP Server is an open-source server providing 50+ tools that cover Kafka, Flink SQL, Schema Registry, Connectors, Tableflow, and more, helping AI assistants manage Kafka clusters, run Flink SQL, and manage schemas and connectors. It works with Confluent Cloud, Confluent Platform, and local Kafka deployments. Compatible with multiple clients like Claude, Cursor, VS Code, and more.

Similar servers

Compare with other MCP servers

Source revision 76cffd5a9644 Data synced 2026-10-12 Read the FMRS scoring method