Command Overview

Batuta provides a unified CLI for the entire transpilation-to-deployment pipeline, plus ML model serving, stack orchestration, and intelligent query interfaces.

Pipeline Commands (5-Phase Workflow)

Command	Phase	Description
`batuta init`	Setup	Initialize project with `batuta.toml`
`batuta analyze`	1	Analyze source codebase (languages, deps, TDG)
`batuta transpile`	2	Transpile source code to Rust
`batuta optimize`	3	MoE backend selection + Cargo profile tuning
`batuta validate`	4	Verify semantic equivalence
`batuta build`	5	Build final binary (release, cross-compile, WASM)

Workflow Management

Command	Description
`batuta status`	Show current workflow phase and progress
`batuta reset`	Reset workflow state to start over
`batuta report`	Generate migration report (HTML/Markdown/JSON)

Intelligence & Query

Command	Description
`batuta oracle`	Knowledge graph queries, RAG search, PMAT code search
`batuta bug-hunter`	Popperian falsification-driven defect discovery
`batuta falsify`	Run Sovereign AI Assurance Protocol checklist

Agent Runtime

Command	Description
`batuta agent`	Autonomous agent runtime (`--features agents`)
`batuta playbook`	Deterministic YAML pipelines with BLAKE3 caching

ML Model Ecosystem

Command	Description
`batuta serve`	Serve models via Realizar (OpenAI-compatible API)
`batuta deploy`	Deploy to Docker, Lambda, K8s, Fly.io, Cloudflare
`batuta mcp`	MCP server for AI tool integration
`batuta hf`	HuggingFace Hub integration

Stack & Data

Command	Description
`batuta stack`	PAIML Stack dependency orchestration
`batuta data`	Data platform integration
`batuta viz`	Visualization frameworks
`batuta content`	Content creation tooling

Global Options

All commands support these flags:

Flag	Description
`-v, --verbose`	Enable verbose output
`-d, --debug`	Enable debug output
`--strict`	Enforce strict drift checking
`--allow-drift`	Allow drift warnings without blocking
`-h, --help`	Print help
`-V, --version`	Print version

Navigate: Table of Contents