Ferrum is an open-source, Rust-based terminal coding agent developed by ominiverdi. Designed as a lightweight, native command-line utility, it provides an interactive environment for AI-assisted programming and automated code generation directly inside the developer's terminal. In daily use, Ferrum runs as an interactive TUI/CLI session featuring session navigation, local instruction management, and loop guard protections to safely manage iterative agent execution. The agent supports automated code inspection and task resolution, with repository-level context handling and support for image-based multimodal prompts alongside standard text instructions. What distinguishes Ferrum is its native Rust implementation, prioritizing low resource overhead, execution speed, and local workflow control. It incorporates built-in visual theming via palette configurations, request-local loop guards to mitigate runaway execution loops, and dedicated benchmarking harnesses to evaluate agent performance. Ferrum is released as an open-source project hosted on GitHub and distributed under standard open-source licensing terms. The project is actively maintained with regular tag releases.
Rust-native implementation — Built entirely in Rust with Cargo packaging for fast startup and minimal binary overhead. - Interactive session navigation — Terminal interface designed for managing multi-turn coding sessions and prompt history. - Built-in loop guards — Safety guardrails that operate request-locally to prevent runaway iteration loops during task execution. - Local agent instructions — Configurable system prompts and instructions kept local to the user repository or environment. - Multimodal prompt support — Handling for image prompts alongside standard terminal text prompts. - Terminal UI themes — Built-in visual palettes seeded on initial run for customized terminal styling. - Integrated benchmarking — Native benchmark suites to evaluate agent reliability and regression integrity.
Use cases
Use Case: Terminal-Based Code Refactoring A developer opens an interactive Ferrum session inside an existing codebase to refactor legacy functions or update internal APIs. The developer supplies instructions while relying on Ferrum's interactive navigation and local instruction files to keep execution scoped.
Use Case: Multimodal UI Implementation A frontend developer passes UI mockups and screenshots directly into Ferrum's prompt workflow. The agent parses the image context alongside the local codebase to generate corresponding UI component code.
Use Case: Safe Scripting and Loop-Guarded Automation A DevOps engineer tasks Ferrum with repetitive repository modifications across several files. Ferrum's loop guards ensure that iterative modifications terminate cleanly without entering unconstrained agent generation cycles.
Who it is for
Ferrum is suited for developers and DevOps engineers who prefer lightweight, native Rust CLI utilities and terminal-centric AI agent workflows. It is ideal for teams wanting local configuration control and safety loop guards, and less suited for developers seeking full GUI/IDE plugins or hosted cloud agent dashboards.
Install guide
Installation Install Ferrum directly from source using Rust's Cargo package manager: bash cargo install --git https://github.com/ominiverdi/ferrum Alternatively, build the binary locally: bash git clone https://github.com/ominiverdi/ferrum.git cd ferrum cargo build --release
First Run Run the executable to seed default palettes and initialize local configuration: bash ferrum Set your required LLM provider API credentials as environment variables before starting an interactive session (for example, standard OPENAI_API_KEY or ANTHROPIC_API_KEY depending on your backend provider configuration).
Example Session Launch an interactive session inside your project directory: bash ferrum Inside the prompt interface, enter your task: text > Analyze src/main.rs and add loop guard error handling for network timeouts.
OpenHands — OpenHands (formerly OpenDevin) is an open-source autonomous AI software development agent and orchestration platform. It provides an extensible runtime environment…
Plandex — Plandex is an open-source terminal-based AI coding agent designed to handle complex, multi-file engineering tasks in large codebases. Created by…
Kiro — Kiro is an agentic engineering platform and CLI coding agent built and operated by AWS. It extends beyond standard code-completion…
Amp — Amp (developed by AmpCode) is an autonomous coding agent platform and remote development harness designed to execute software engineering tasks…
OpenCode — OpenCode is an open-source AI coding agent developed by Anomaly that operates across terminal environments, desktop applications, and IDE extensions.…
Ona — Ona (formerly Gitpod's agentic platform, now part of OpenAI) is an enterprise-grade cloud agent orchestration platform that manages and runs…