Connector Capability Matrix

This page is the quick-reference matrix for engineers deciding which AxonFlow MCP connectors to use in a real system. It complements the individual connector guides by putting the practical questions in one place:

• Which connector types exist in the current codebase?
• Which ones are available in community builds versus enterprise builds?
• Which auth model does each connector usually need?
• Which connectors are best for local development, evaluation, or broader enterprise rollout?

If you are planning a serious multi-agent or governed retrieval system, this page is usually faster than opening ten separate connector pages first.

How To Read The Matrix

• Edition refers to current build-path availability in the runtime.
• Auth model is the most common credential pattern, not every possible option.
• Runtime fit is the typical starting environment where the connector makes operational sense.
• Capabilities summarizes the high-level operations the connector exposes through MCP.

Connector Matrix

Connector type	Category	Edition	Typical auth model	Runtime fit	Common capabilities
postgres	Database	Community	connection string or user/password	local, self-hosted, evaluation, enterprise	query, execute
mysql	Database	Community	connection string or user/password	local, self-hosted, evaluation, enterprise	query, execute
mongodb	Database	Community	Mongo URI	local, self-hosted, evaluation, enterprise	query, execute
cassandra	Database	Community	host/user/password	evaluation, enterprise	query, execute
redis	Cache / data	Community	URI or host/password	local, self-hosted, evaluation, enterprise	query, execute
http	API	Community	bearer token, API key, or custom headers	local, evaluation, enterprise	query, execute
s3	Object storage	Community	IAM role, access keys	evaluation, enterprise, AWS-heavy teams	query, execute
azure_blob	Object storage	Community	account key, SAS, managed identity	evaluation, enterprise, Azure estates	query, execute
gcs	Object storage	Community	service account or workload identity	evaluation, enterprise, GCP estates	query, execute
amadeus	SaaS / travel	Enterprise	OAuth client credentials	enterprise	query, execute
salesforce	SaaS / CRM	Enterprise	OAuth	enterprise	query, execute
slack	SaaS / collaboration	Enterprise	bot token / app token	enterprise	query, execute
snowflake	Data warehouse	Enterprise	user/password, key pair, or DSN	enterprise	query, execute
hubspot	SaaS / CRM	Enterprise	OAuth or private app token	enterprise	query, execute
jira	SaaS / ticketing	Enterprise	PAT or OAuth	enterprise	query, execute
servicenow	SaaS / ITSM	Enterprise	user/password or OAuth	enterprise	query, execute

What The Current Runtime Actually Registers

On current main, the community connector factory registers:

• postgres
• mysql
• mongodb
• cassandra
• redis
• http
• s3
• azure_blob
• gcs

Enterprise build paths add:

• amadeus
• salesforce
• slack
• snowflake
• hubspot
• jira
• servicenow

That means object-storage connectors are no longer just an enterprise teaser in the runtime itself. They are part of the current community connector estate and should be considered first-class options for governed document and file workflows.

Choosing The Right Connector Set

If you are validating AxonFlow quickly

Start with:

• postgres or mysql for structured data
• http for internal APIs
• s3 if your workflow revolves around documents or generated artifacts

This gives you enough surface area to prove policy enforcement, response redaction, and auditability without taking on SaaS admin complexity too early.

If you are preparing for a broader rollout

The signal that you are moving beyond a simple community setup is usually one of:

• more than one business system has to be governed
• files and records need the same policy posture
• the platform becomes shared across multiple teams
• procurement and identity constraints matter as much as raw connectivity

That is where Community vs Evaluation vs Enterprise and Community To Enterprise Migration become more important than the individual connector page alone.

Operational Advice

• Use Runtime Configuration when you want to understand DB, file, and env precedence.
• Use Response Redaction when the connector will touch sensitive data.
• Use Policy Enforcement when external orchestrators need pre-check or post-check behavior around tool calls.

Related Docs

• MCP Connectors Overview
• Runtime Configuration
• Policy Enforcement
• Community vs Evaluation vs Enterprise