<img src="assets/rustchain_logo_tight_transparent.png" alt="RustChain" width="400"> Keep your existing workflows. RustChain runs them all. <a href="https://github.com/Michael-A-Kuykendall/rustchain/actions/workflows/ci.yml"><img src="https://github.com/Michael-A-Kuykendall/rustchain/actions/workflows/ci.yml/badge.svg" alt="CI"></a> <a href="https://opensource.org/licenses/MIT"><img src="https://img.shields.io/badge/License-MIT%20OR%20Apache--2.0-blue.svg" alt="License: MIT OR Apache-2.0"></a> <a href="https://www.rust-lang.org"><img src="https://img.shields.io/badge/rust-1.70+-orange.svg" alt="Rust"></a>

RustChain consumes workflows from the tools you already use:

| Platform | Status | Command | |----------|--------|---------| | LangChain (Python) | ✅ Supported | rustchain transpile lang-chain script.py | | Apache Airflow | ✅ Supported | rustchain transpile airflow dag.py | | GitHub Actions | ✅ Supported | rustchain transpile github-actions workflow.yml | | Kubernetes | ✅ Supported | rustchain transpile kubernetes deployment.yaml | | Docker Compose | ✅ Supported | rustchain transpile docker-compose compose.yaml |

You don't have to rewrite anything. Point RustChain at your existing files and it generates executable missions.

# Your existing LangChain script
rustchain transpile lang-chain my_agent.py -o mission.yaml

# Now run it with RustChain
rustchain run mission.yaml

Why migrate?

| Problem with current tools | RustChain solution | |---------------------------|-------------------| | Python's GIL limits parallelism | True multi-threading, no GIL | | GC pauses cause latency spikes | Deterministic memory management | | Container startup overhead | Native binary, instant startup | | Vendor lock-in | Universal format, portable everywhere |

How Transpilation Works

RustChain's transpiler parses your existing workflow definitions and generates equivalent RustChain missions:

Your Workflow          RustChain              Output
─────────────          ─────────              ──────
LangChain.py      ──▶  Transpiler        ──▶  mission.yaml
Airflow DAG            (parses & converts)    (executable)
GitHub Actions
K8s manifest
Docker Compose

What gets converted

Steps/Tasks → Mission steps with proper types
Dependencies → depends_on relationships preserved
Configuration → Parameters mapped to RustChain equivalents
Secrets/Env vars → Environment variable references preserved

Auto-detection

Don't know the format? RustChain figures it out:

rustchain transpile auto my_workflow.py
# Detects: LangChain Python
# Output: my_workflow.yaml

Supported conversions

| From | Detected by | What's preserved | |------|------------|------------------| | LangChain | from langchain, from openai | Chains, agents, tools, prompts | | Airflow | @dag, DAG(, airflow imports | Operators, dependencies, schedules | | GitHub Actions | on:, jobs: | Steps, runners, secrets | | Kubernetes | apiVersion:, kind: | Containers, resources, volumes | | Docker Compose | services:, image: | Services, networks, volumes |

Quick Start

1. Install

git clone https://github.com/Michael-A-Kuykendall/rustchain
cd rustchain
cargo build --release --features "cli,tools,llm"

2. Convert an existing workflow

# Have a LangChain script?
rustchain transpile lang-chain your_script.py -o mission.yaml

# Have an Airflow DAG?
rustchain transpile airflow your_dag.py -o mission.yaml

# Not sure what format?
rustchain transpile auto your_file.py -o mission.yaml

3. Validate and run

# Check it first
rustchain mission validate mission.yaml

# Dry run (no side effects)
rustchain run mission.yaml --dry-run

# Execute
rustchain run mission.yaml

Native Mission Format

If you want to write missions directly (instead of converting), use YAML:

name: "Data Pipeline"
version: "1.0"
steps:
  - id: "fetch_data"
    name: "Fetch from API"
    step_type: "http_request"
    parameters:
      url: "https://api.example.com/data"
      method: "GET"

  - id: "process"
    name: "Process with LLM"
    step_type: "llm"
    depends_on: ["fetch_data"]
    parameters:
      provider: "ollama"
      model: "llama2"
      prompt: "Summarize: ${fetch_data.output}"

  - id: "save"
    name: "Save results"
    step_type: "create_file"
    depends_on: ["process"]
    parameters:
      path: "output.txt"
      content: "${process.output}"

CLI Reference

rustchain run <file>           Execute a mission
rustchain transpile <cmd>      Convert from other formats
rustchain mission <cmd>        Mission management
rustchain safety <cmd>         Safety validation
rustchain tools <cmd>          Tool management
rustchain audit <cmd>          Audit queries
rustchain policy <cmd>         Policy management
rustchain config <cmd>         Configuration
rustchain interactive          Interactive mode

See rustchain --help for full details.

Documentation

Requirements

Rust 1.70+: Required for compilation
Optional: Ollama or compatible LLM backend for AI features

License

Dual-licensed under MIT or Apache-2.0, at your option.

See LICENSE-MIT and LICENSE-APACHE.

Contributing

RustChain is open source but not open contribution. See CONTRIBUTING.md.

Security

To report vulnerabilities, see SECURITY.md.

Rustchain

Install / Use

README