Lightweight • Local-First • Powerful
Explore, transform, and visualize CSV/TSV data without leaving VS Code
VizFlow brings data exploration and lightweight analytics directly into your editor. No need to switch contexts, open external tools, or worry about data leaving your machine.
- 📊 Immediate dataset profiling — understand your data at a glance
- 🔄 Visual, repeatable transforms — preview before applying changes
- 🗄️ RBQL-powered SQL-like querying — familiar syntax, instant results
- 📈 Interactive charts — bar, line, scatter, and pie visualizations
- 🎯 Local-first by default — your data stays on your machine
- Column types, null counts, distinct values, top values
- Quick stats (sum, avg, min, max, count) via the Dataset Summary Dashboard, or as a repeatable pipeline step with the Workflow Builder's Aggregate / Group By activities
- Find duplicates in your dataset
- List distinct values for any column
- Build queues of transformation rules
- Preview first rows before applying changes
- Apply & save with confidence
- Write SQL-like queries using
a1,a2or header names, with syntax-highlighted keywords as you type - JOIN support across two files, right in the query editor
- Pre-defined query templates to get started fast (filter, sort, aggregate, join, and more)
- History tracking for reproducible analysis
- Progress indicators for long-running queries
- Bar charts, line charts, scatter plots, pie charts
- Live mapping with instant updates
- Value-based set comparison across two files
- CSV / JSON export
- DuckDB-compatible SQL with
{{INPUT_PATH}}/{{OUTPUT_PATH}}placeholders
- Workflow parameters: declare inputs (name, label, type, required, default) that flow into any step config as
{{paramName}} - Call Workflow: run another
.vizflowfile as a sub-workflow, pass parameters, and receive its output dataset and variables back - Circular-call detection and a max call-depth guard keep nested workflows safe
- Edit parameters and Call Workflow mappings with the built-in key/value editors (🧩)
- Schedule
.vizflowworkflows on a cron schedule or as a one-time run at a future date/time - Built-in variables (
{{timestamp}},{{date}},{{time}},{{workflowName}}, …) and per-job parameters are injected into every scheduled run - Optional timezone, watch folder (trigger on new files, with a filter like
*.csv), and webhook / email notifications - Jobs persist in VS Code's global storage and auto-start with the extension, so schedules run even if the Scheduler panel is never opened
- Edit jobs in place, pause/resume/stop, and browse past runs from the panel
- Connect to MongoDB, MySQL, and PostgreSQL from the VizFlow: Data Sources panel
- Credentials (passwords / connection strings) are stored in your OS keychain via VS Code SecretStorage — workflows reference connections by friendly name and stay safe to commit to git
- Visual query builder: pick a collection/table, tick columns, add filters (equals / contains / greater than / between / …), sort, and preview rows before loading
- Advanced tab for raw SQL
WHEREclauses or Mongo filter JSON — all reads are strictly read-only - "Add to Workflow Builder" generates a ready-to-run
readMongo/readSqlactivity that pipes the data into your pipeline
- Install VizFlow from the VS Code Marketplace
- Open a CSV/TSV file in the editor
- Press
Ctrl+Shift+P(Cmd+Shift+Pon macOS) - Run any
VizFlow:command, for example:VizFlow: Dataset Summary Dashboard
| Category | Command |
|---|---|
| Profiling & Quality | Show Statistics, Find Duplicates, List Distinct Values |
| Transformation | Transform Column (CLI), Transform Column (Visual) |
| Analysis | Dataset Summary Dashboard, Interactive Charts |
| Comparison | Compare CSV Files |
| RBQL | Run RBQL Queries (with syntax highlighting, JOINs, and templates), Export Results |
| Workflow Builder | Aggregations (Sum / Average / Min / Max / Count) and Group By, as pipeline activities |
| Scheduler | Workflow Scheduler (add, run, pause/resume, stop, and view history for jobs, all in one panel) |
| Data Sources | External Data Sources (MongoDB / MySQL / PostgreSQL) |
| About | View Extension Author Info |
- Use positional references (
a1,a2, ...) or enable Has Header Row to use header names - Export results to CSV, JSON, or DuckDB SQL scripts
- Placeholders like
{{INPUT_PATH}}/{{OUTPUT_PATH}}make scripts reproducible
VizFlow runs locally by default — parsing, transforms, and queries are performed on your machine. Any cloud or AI features would be opt-in and clearly indicated.
- Workflow Builder Guide: docs/workflow-builder-guide.md — complete user guide with activity reference, real-world use cases, and troubleshooting
- Activity Catalog: docs/workflow-catalog.md — auto-generated reference for every activity type and config field
- Architecture & Implementation: ARCHITECTURE.md
- AI-assisted workflow authoring:
AI_CONTEXT.md,docs/workflow-catalog.mdanddocs/workflow-schema.jsonlive in the GitHub repository (dev-facing; not shipped in the VSIX)
Search VizFlow in the Extensions view and install.
git clone https://github.com/ESAditya1729/VizFlow.git
cd VizFlow
npm install
# Press F5 in VS Code to launch the Extension Development HostRun these locally during development:
npm install # Install dependencies
npm run lint # Run linter
npm test # Run tests- GitHub: ESAditya1729/VizFlow
- Creator: Aditya Mukherjee (IBM)
- LinkedIn: Aditya Mukherjee
MIT — see LICENSE
Built with ❤️ by Aditya Mukherjee · IBM