wreckit

wreckit is an open-source CLI agent harness and orchestration framework created by Mike Hostetler. Built using TypeScript and designed around automated development workflows, it provides a structured environment for managing and executing AI-driven codebase modifications and specification workflows. In standard use, wreckit operates directly from the terminal to execute automated code improvements, bug fixes, and specification-driven development cycles. It leverages dedicated project configurations stored in directories like `.wreckit` to orchestrate tasks across repositories, integrating with underlying LLM backends and agents defined in project specification files. What distinguishes wreckit is its cybernetic system approach and focus on structured, spec-driven execution loops. By decoupling harness orchestration logic from standard single-prompt CLI assistants, it coordinates agent tasks through defined specifications, fixes, and repeatable run scripts. wreckit is open-source software distributed on GitHub under the repository's open license, managed with modern JavaScript/TypeScript package managers like Bun and npm.

Tags: cli, codebase, developer tools, experiments, OpenSource

Visit wreckit

Key features

• Specification-driven execution — Orchestrates agent tasks based on structured markdown specifications and configuration files. - Project-level workspace configuration — Uses dedicated .wreckit directories to define project parameters, agent rules, and execution context. - Multi-runtime compatibility — Built for TypeScript runtimes with support for Bun and Node.js package managers. - Agent harness architecture — Manages execution loops, fixes, and script pipelines separately from the underlying model logic. - Standardized agent documentation support — Integrates repository-level rule files such as AGENTS.md and CLAUDE.md to guide agent actions.

Use cases

Use Case: Automated Specification Execution Developers use wreckit to process structured project specifications and execute multi-step codebase updates without manually prompting the model for each individual file modification.

Use Case: Coordinated Codebase Fixes Teams apply automated bug fixes and maintenance patches across a repository by defining targeted fix scripts and execution profiles in the .wreckit workspace.

Use Case: Agent Context Management Engineers leverage wreckit to standardize AI agent behavior and workspace boundaries across distributed projects using version-controlled harness configurations.

Who it is for

wreckit is designed for software engineers and maintainers seeking an automated, spec-driven harness to orchestrate AI agent tasks in terminal and CI environments. It is best suited for teams working in TypeScript/JavaScript ecosystems who want reproducible agent workflows rather than casual conversational chat interfaces.

Install guide

Installation Install wreckit locally or clone the repository to run with Bun or Node.js: bash git clone https://github.com/mikehostetler/wreckit.git cd wreckit bun install Or using npm: bash npm install

First Run Ensure appropriate environment variables and API credentials for your model provider are configured in your shell environment: bash export ANTHROPIC_API_KEY="your-api-key" Initialize or verify the project configuration within your repository: bash ls -la .wreckit

Example Session Run the harness against your project specifications: bash bun run ./bin/wreckit --spec specs/example-feature.md

Part of CLI Coding Agents

Categories: AI Coding Agents, Developer Tools, Experiments

Related CLI agents

• TeDDy — TeDDy (developed by atte500) is an open-source test-driven execution harness and orchestrator written in Python. It is designed to manage…
• Continue — Continue is an open-source AI code assistant and agent framework originally developed by the Continue team (continuedev) and recently acquired…
• Grite — Grite is an open-source terminal coding agent harness developed by Neul Labs. Implemented in Rust, it provides a lightweight command-line…
• wt — wt is an open-source CLI harness and workspace orchestration tool maintained by Sahil Ahuja on GitHub. It is built to…
• uacos — UACOS (by developer caotiensinh) is a local-first code intelligence, context-compression, and release-gate toolkit designed to operate as an agent harness…
• Agent Memory System — Agent Memory System is an open-source agent orchestration harness developed by RavByte-AI. It provides structured memory retention and retrieval mechanisms…