Remote Factory is an open-source orchestration harness and agent framework maintained by akashgit on GitHub. Designed to standardize and coordinate AI coding skills, agent personas, and background tasks, it bridges local development workflows with automated agent execution environments across supported model platforms. Day-to-day operation revolves around CLI commands like `factory` alongside plugin integrations for Claude and OpenAI Codex environments. Remote Factory manages modular agent roles (such as Failure Analysts and specialized coding sub-agents), routes commands automatically across task modes, and supports cross-project code analysis using dedicated slash commands and skills directories. What sets Remote Factory apart is its dual integration design, featuring native plugin configurations for both `.claude-plugin` and `.codex-plugin` architectures, combined with a containerized runtime (`factory contained setup`) for sandboxed task execution. This structure allows teams to deploy consistent AI agent skills across different vendor interfaces without duplicating agent definitions. Remote Factory is an open-source project distributed under standard repository licensing, packaged via Python (`pyproject.toml`) and distributed with prebuilt runtime container images on GitHub Packages (GHCR).
Tags: agent harness, ai agent, cli, coding agent, developer tools, OpenSource
.claude-plugin) and Codex (.codex-plugin) environments. - Containerized Agent Runtime — Offers isolated execution via pre-built GHCR runtime images initialized with factory contained setup. - Specialized Agent Roles — Includes structured agent personas, including Failure Analyst and Refiner roles, coordinated automatically for debugging and coding. - Automated Task Routing — Automatically routes development commands and prompts to the appropriate sub-agent or execution mode. - Cross-Project Insights — Supports multi-repository inspection and skill harvesting through commands like /factory:study.factory contained setup, they spin up a pre-configured containerized runtime to let agent personas generate and test code in isolation./factory:study --projects-dir to aggregate insights, analyze conventions, and feed project knowledge back into standard agent skill definitions.Remote Factory is suited for Python developers, DevOps engineers, and teams wanting a containerized orchestrator that operates across both Claude and Codex plugin ecosystems. It is less suitable for users looking for a zero-configuration GUI or basic single-model chat wrappers.
pip / pipx: bash git clone https://github.com/akashgit/remote-factory.git cd remote-factory pip install .bash factory contained setup Configure required LLM provider API credentials in your environment: bash export ANTHROPIC_API_KEY="your-anthropic-api-key" export OPENAI_API_KEY="your-openai-api-key"bash # Run cross-project study and agent skill discovery factory study --projects-dir ./workspace # Run containerized task execution factory run "Refactor database access layer to use async connection pools"Part of CLI Coding Agents
Categories: AI Coding Agents, Code Assistants, Developer Tools