ToolVault

ToolVault is a portable, self-contained service that delivers curated collections of analysis tools for scientists, analysts, and developers. It provides an interactive browser-based interface for discovering, running, and understanding tools — complete with git history, legacy version access, and spatial output visualisation.

Key Features

• Curated Tool Collections – Tools packaged with all their dependencies in a zip bundle.
• Self-Describing – Each bundle contains metadata (index.json) describing the tools, inputs, outputs, and history.
• Interactive UI – Browse tools, configure inputs, run them, and view outputs directly in the browser.
• Version Awareness – Access and run specific historical versions of a tool.
• Spatial Rendering – LeafletJS visualisation for spatial outputs (e.g., GeoJSON FeatureCollections).
• Jupyter Integration – Embedded notebook server for viewing and running demo notebooks for each tool.
• Multiple Delivery Modes – Runs as a standalone Flask/FastAPI server or as a VS Code extension.

Current Phase

We are in the early UI mockup phase, focusing on:

1. Building a bare static UI driven by index.json.
2. Adding mock tool calls to simulate execution.
3. Implementing the full browsing UI with output rendering (including LeafletJS).

Planned Roadmap

1. Phase 1 – UI Mockup

   • Bare static UI from index.json (browse tools, view details, inputs/outputs, run button).
   • Mock tool calls with placeholder outputs.
   • Full browsing UI with LeafletJS output rendering and git history view placeholder.

2. Phase 2 – Indexer Development

   • ToolVault indexer to scan a Git repo of tools, generate index.json and metadata.
   • Bundle dependencies, notebooks, and metadata into a zip for offline use.

3. Phase 3 – Backend Service

   • Flask/FastAPI service that loads bundles and exposes tools via a REST API.
   • Embedded Jupyter server for running demo notebooks.

4. Phase 4 – VS Code Integration

   • ToolVault available as a VS Code extension, integrated into the Debrief UI.

Repository Structure (proposed)

/client         # SPA frontend
/server         # Flask/FastAPI backend
/indexer        # Bundle creation scripts
/docs           # Documentation and design notes
/examples       # Example tool bundles
README.md

License

TBD – likely permissive open source license to encourage adoption.