live wire
▸Snowflake’s Cortex Agent quickstart routes metrics and transcripts with prompt rules—but documents no authorization boundarySnowflake▸Oracle Select AI generates SQL/PGQ for property graphs—but errors if one AI profile mixes graphs with tables or viewsOracle▸Domo’s data team standardizes MCP builds with shared scaffolding; beta toolkits inherit user permissionsDomo▸Databricks proposes five outcome signals for agentic BI—and says dashboard counts and license usage miss decision qualityDatabricks▸One vibe-coded database app needed 56+ human test hours and three repair rounds despite AI Studio’s verification claimsMancas & Mancas / arXiv▸BigQuery AI.IF’s default cost mode trains a local model for most rows—and drops row-error thresholdsGoogle Cloud▸AWS’s multi-tenant data-agent pattern gives the broker Lambda no data permissions; Cognito roles carry Lake Formation scopeAWS Big Data Blog▸Amazon OpenSearch’s optimized engine adds nested agent traces, but AWS reserves it for log analytics—not vector searchAWS▸Two interpretation-based provenance semantics cannot express SQL bag difference, new theorem showsKolaitis / arXiv▸Databricks Feature Views reuse one definition for training and live scoring, but online lookup still returns latest available stateDatabricks▸Databricks Genie’s Slack and Teams previews default channel answers to public, even when viewers lack source-data accessDatabricks▸Snowflake’s two-tool GraphRAG agent beat a seven-tool graph setup 70% to 60% on 22 ontology questionsSnowflake Engineering▸Semantix opens an Apache-2.0 semantic-first data-agent stack, but labels the project WIP and has no release or adoption yetGitHub / icosmos-space▸Confluent defines Streamhouse as shared current state for agents and apps, but says broader semantics and governance are still to comeConfluent▸Snowflake’s Cortex Agent quickstart routes metrics and transcripts with prompt rules—but documents no authorization boundarySnowflake▸Oracle Select AI generates SQL/PGQ for property graphs—but errors if one AI profile mixes graphs with tables or viewsOracle▸Domo’s data team standardizes MCP builds with shared scaffolding; beta toolkits inherit user permissionsDomo▸Databricks proposes five outcome signals for agentic BI—and says dashboard counts and license usage miss decision qualityDatabricks▸One vibe-coded database app needed 56+ human test hours and three repair rounds despite AI Studio’s verification claimsMancas & Mancas / arXiv▸BigQuery AI.IF’s default cost mode trains a local model for most rows—and drops row-error thresholdsGoogle Cloud▸AWS’s multi-tenant data-agent pattern gives the broker Lambda no data permissions; Cognito roles carry Lake Formation scopeAWS Big Data Blog▸Amazon OpenSearch’s optimized engine adds nested agent traces, but AWS reserves it for log analytics—not vector searchAWS▸Two interpretation-based provenance semantics cannot express SQL bag difference, new theorem showsKolaitis / arXiv▸Databricks Feature Views reuse one definition for training and live scoring, but online lookup still returns latest available stateDatabricks▸Databricks Genie’s Slack and Teams previews default channel answers to public, even when viewers lack source-data accessDatabricks▸Snowflake’s two-tool GraphRAG agent beat a seven-tool graph setup 70% to 60% on 22 ontology questionsSnowflake Engineering▸Semantix opens an Apache-2.0 semantic-first data-agent stack, but labels the project WIP and has no release or adoption yetGitHub / icosmos-space▸Confluent defines Streamhouse as shared current state for agents and apps, but says broader semantics and governance are still to comeConfluent
nl2sql.ai
releaseRelease analysis

CONNX puts legacy databases behind MCP—but “read access” still needs a production boundary

Version 15.0 packages OAuth, optional Vault-backed source credentials and SQL guidance into a containerized MCP server. Teams should verify what the Community Edition actually prevents before connecting an agent.

By The News Desk· Oct 18, 2026the quick take — two AI hosts go live when you do

Software AG’s October release gives enterprise teams a direct way to put CONNX-connected data behind an AI-tool interface. CONNX 15.0 adds an MCP Server Community Edition that can run as a container and expose remotely accessible MCP over streamable HTTP or HTTPS.

That matters because CONNX is positioned around access to legacy, relational and other data sources, plus virtualization and replication. The MCP server therefore is not merely another chatbot connector: it can become a new route from an agent to systems that were previously reached through desktop tools, dashboards or integration pipelines.

The release separates two authentication hops

The release highlights list OAuth between the MCP client and the CONNX MCP server. Authentication from CONNX to the underlying data sources can use Vault, but that integration is optional.

Operators should model those as separate trust boundaries. OAuth answers which client may call the MCP endpoint. Source credentials determine what the resulting session can do inside each connected database. A successful OAuth check does not, by itself, prove least privilege at the database hop.

“Read access” is the starting claim, not the finished control

Software AG describes the Community Edition as providing read access to CONNX data sources. The announcement does not spell out whether that guarantee is enforced by the MCP tool surface, the CONNX layer, source-account grants, SQL parsing, or some combination of them. It also does not publish limits for row counts, query cost, timeouts or concurrent requests.

Before production use, teams should test the boundary with the actual source identity: attempt writes and DDL, invoke stored procedures, query sensitive system catalogs, and run deliberately expensive reads. Database-native read-only grants and workload controls should remain the backstop until the vendor documents stronger enforcement.

The SQL skill needs its own regression set

The package also includes a “CONNX SQL expert skill” intended to improve an LLM’s use of CONNX SQL. That is useful for dialect guidance, but it introduces another versioned behavior surface. Teams should keep a small golden set covering joins, date handling, null behavior and source-specific syntax, then rerun it whenever the server image or skill changes.

The practical rollout sequence is straightforward: pin the container image; require OAuth; prefer Vault-backed source credentials; give those credentials database-native read-only rights; cap query cost and duration; and log the client identity, source identity and executed SQL together. CONNX 15.0 supplies the connection path. Production safety still depends on proving what happens at both ends of it.

sources

Filed by The News Desk. Corrections: desk@nl2sql.ai · Our standards →

comments · 0

    Comments are moderated before they appear. Your email is used once to confirm it is you — never shown, never sold. Corrections and questions get an answer from the desk when we have one.

    more from the desk