iCloud sync for saved queries and pipelines

Optional iCloud-backed sync of your QueryFlow workspace across multiple Macs. Pipelines, saved queries, and schedules sync automatically. Credentials stay in each Mac's Keychain — never synced.

SSH tunneling for Postgres and MySQL

Connect to databases behind firewalls via SSH tunnel — the standard pattern for production databases requiring jump host access. Supports password and key-based SSH authentication, multi-hop tunnels.

Custom Python package install in Flow Books

Currently Flow Books ship with pandas, numpy, matplotlib, requests, boto3 pre-installed. Custom pip install via Pyodide would let users add their own dependencies per workspace (e.g., openai, anthropic SDKs, custom ML packages).

Collaboration via shared queries and pipelines

Optional sharing of queries and pipelines with team members. Shareable links generate a read-only or editable view. Comments inline on specific cells. Designed for small teams where iCloud isn't enough but full enterprise collab is overkill.

AI chart suggestions in Flow Books

After running a query, Claude AI analyzes the result columns and suggests the best chart type. One click to generate the matplotlib/plotly code in a new Python cell. Smart defaults based on data types and column count.

Real-time CDC mode for Postgres and MySQL

Stream change events from Postgres logical replication or MySQL binlog instead of running scheduled snapshots. Sub-second replication SLAs for use cases that need them. Trade-off: meaningfully more complex setup than batch snapshots.

Query version history with lightweight snapshots

Automatic versioning of saved queries with diff view. See what changed between versions. Restore any previous version with one click. Lightweight SwiftData snapshots — no Git complexity.

Export Flow Books to Notion and Google Sheets

Publish your Flow Book results directly to a Notion page or Google Sheets workbook. Cells render as proper Notion blocks (code blocks, tables, charts as images). Useful for sharing analysis with non-QueryFlow team members.

HubSpot connector (source and destination)

Read Contacts, Companies, Deals, Tickets from HubSpot. Write back records for marketing list imports and reverse ETL. OAuth 2.0 with full API scope.

Calendar-based job triggers with iCal/iCloud sync

Schedule jobs to fire relative to calendar events. Examples: '48 hours before every home game' for sports analytics, '1 hour after every standup' for team operations. iCal import plus live iCloud Calendar sync with event-matching rules.

SQLite as a bundled connector

SQLite support via GRDB.swift. Bundled directly into QueryFlow with no separate driver install. Useful for local development, sample data, and quick file-based projects.

MongoDB connector

MongoDB support via mongodb-nio or REST API. Document-aware schema explorer that handles MongoDB's flexible schema. Aggregation pipeline support in the editor.

Salesforce as a write destination beyond field mapper

Extending Salesforce write operations beyond the Visual ETL field mapper — direct destination support in pipelines so any source can push to Salesforce without the mapper UX.

ClickHouse connector

ClickHouse via the HTTP interface. Fast columnar query results with ClickHouse-specific syntax highlighting and EXPLAIN support.

DynamoDB connector

Amazon DynamoDB support via the AWS SDK. Table and item-level operations. PartiQL query support for relational-style access to DynamoDB tables.

Show 'N of M rows' when results are capped

Make truncation visible in the results grid and exports.

CSV import: quoted newlines and escaped quotes

Google Sheets: update and clear ranges, rate-limit backoff

Databricks connector

Native Databricks support via the SQL Statement Execution API. OAuth and PAT authentication. Works with both Unity Catalog and legacy workspaces. Lakehouse-aware schema explorer.

BigQuery connector

Native Google BigQuery support alongside Snowflake and Redshift. Service account authentication, standard SQL dialect, scheduled queries with cost-aware billing tier selection.

Excel .xlsx as a data source

Currently CSV/TSV files work as sources. Extending the SpreadsheetStrategy to also handle Excel .xlsx workbooks — read any tab, handle multiple sheets, type inference for date/number columns.

Claude writes the query — it knows every table

Claude AI suggests SQL completions as you type, using full awareness of your connected database's schema. Tab to accept the suggestion. Works against any connected database — Snowflake, Redshift, Postgres, MySQL, Google Sheets, Salesforce.

ETL migration from legacy tools — the kill shot

Connect QueryFlow to your existing Informatica, AWS Glue, Airflow, or Talend instance. Claude reads every job definition and rebuilds it in QueryFlow's native format. The goal: cancel your enterprise ETL contract in an afternoon.

Salesforce reverse ETL with OAuth and field mapper

Push data into Salesforce from any source (Snowflake, Redshift, Postgres, MySQL, CSV). OAuth 2.0 authentication. Insert, Update, Upsert sync modes. Maps source columns to Salesforce field names with AI assistance.

Slack webhook destination

Push query results or job status updates to Slack channels via incoming webhooks. Formatted tables using Slack Block Kit. Conditional alerts based on result thresholds (e.g., post only when at-risk deals exceed N).

Flow Books — SQL + Python notebooks with shared state

Mixed SQL and Python cells in one notebook. Query results auto-inject as a pandas DataFrame named df. Pyodide WASM runtime ships pandas, numpy, matplotlib pre-installed. Markdown cells for documentation.

Google Sheets as a write destination

Currently Google Sheets is source-only. Extending the destination model so pipelines can write results back to a Google Sheets tab. Includes append mode and replace mode.

Visual ETL pipeline with bezier-curve field mapper

Drag-and-drop pipeline builder with curved connection lines between source and destination fields. Auto-map algorithm matches columns by name similarity. Claude AI 'Suggest Mappings' button for harder cases. Transform chips for UPPERCASE, lowercase, TRIM, date format conversion, custom Python.

Scheduled jobs that run while your Mac sleeps

Cron, interval, daily, weekly, and manual triggers. Jobs run via SMAppService (App Store compliant login item). Missed jobs catch up automatically on next wake. Observatory dashboard tracks every run with full history and logs.

Claude AI is now error-aware

Claude's chat panel now sees your last 5 query errors automatically, including timestamps and SQL snippets. When a query fails, ask Claude what went wrong and it analyzes the actual error message in context of your schema. The red ErrorContextChip surfaces in the chat when recent errors are present.

Google Sheets connector with live OAuth

Connect to Google Sheets via OAuth 2.0 through your system browser. Pulls live data on every query — your Sheets edits show up immediately in QueryFlow. Sheets API v4 with token refresh handled automatically.

Webhook destination for custom HTTP POST

Generic HTTP POST destination for any service that accepts JSON webhooks. Configure URL, headers, and payload format. Works with Microsoft Teams, Discord, Zapier, and any custom endpoint.