cofactor-vigil 0.1.2

Vigil CLI: Observable, collaborative, reproducible science platform

Vigil CLI

Observable, collaborative, reproducible science platform.

Installation

Base Installation

# Install core CLI with uv
uv tool install cofactor-vigil

# Or with pip
pip install cofactor-vigil

The base installation includes only the core CLI commands (new, dev, build, run, promote, anchor, doctor, etc.).

MCP Server Installation

To use the MCP server (vigil mcp serve), install with the mcp extras:

# Install with MCP server support
uv tool install "cofactor-vigil[mcp]"

# Or with pip
pip install "cofactor-vigil[mcp]"

The mcp extras include:

• mcp>=1.1.0 - MCP protocol server
• polars>=0.20.0 - Fast data frame library for data previews

Quick Start

# Create a new project from a template
vigil new imaging-starter my-project
cd my-project

# Install dependencies
uv sync

# Preview the pipeline (dry-run)
vigil dev

# Execute the pipeline
vigil run --cores 4

# Promote artifacts to receipts
vigil promote

# Anchor receipts with Merkle proofs
vigil anchor

Commands

Project Management

• vigil new <template> [path] - Create a new project from a template
• vigil new --list - List available templates

Development

• vigil dev - Dry-run the pipeline and preview the DAG
• vigil build - Execute the pipeline without promotion
• vigil run - Execute targets and optionally promote
• vigil conformance - Check outputs against golden baselines

Receipt Management

• vigil promote - Generate Vigil receipts from artifacts
• vigil anchor - Create Merkle anchors for receipts
• vigil url - Print the vigil:// URL for the project

Health & Maintenance

• vigil doctor - Run repository health checks
• vigil spec sync - Sync workspace.spec.json with vigil.yaml
• vigil spec --dry-run - Preview workspace spec changes

Workbench & AI

• vigil ui bootstrap - Generate Workbench configuration
• vigil mcp serve - Start the MCP server for assistants
• vigil ai propose - Generate auto-target suggestions
• vigil ai apply - Execute auto-target proposal

Templates

Available starter templates:

• imaging-starter: Full-featured imaging pipeline with sciencecast timeline
• minimal-starter: Smallest viable template for getting started

Documentation

• Getting Started
• CLI Reference
• Creating Custom Templates

License

Apache-2.0