deepseek-ai/deepseek-harness
DeepSeek Harness (dsh) is an open-source agent harness from DeepSeek AI built on an everything-is-a-plugin architecture over the Cordis framework, started from npm with a local web UI.
Awesome AI Agents › AI Developer Tools
Command Code is a terminal coding agent distributed as the command-code npm package and started with the cmd command inside a project directory. The README positions it as a frontier coding agent that both builds software and continuously learns the coding taste of the person using it: it ships full-stack projects, adds features, fixes bugs, writes tests and refactors existing code, while every accept, reject and edit becomes a signal that shapes a taste profile. The taste architecture is built around a model the project calls taste-1, described as meta neuro-symbolic AI that enforces the invisible logic behind coding choices rather than a static list of rules. Taste profiles are portable and can be shared across a team with npx taste push and npx taste pull, and the README argues that explicit rules decay over time while learned taste compounds. The agent runs an interactive session mode with keyboard shortcuts and input modes: typing a slash opens the command menu, an exclamation mark switches to Bash mode, and an at sign triggers file path mention autocomplete. Documentation hosted on the project website covers a quickstart onboarding flow, core concepts including interactive mode, a CLI reference listing flags and subcommands, a description of how taste works, and a set of workflow recipes that each present a concrete scenario, the prompts to use and what to expect. Feedback goes through a /feedback command or GitHub issues, and there is a Discord community and a launch announcement describing the correction-loop problem the taste model is meant to address. The repository README is largely a pointer to the hosted documentation, so implementation details such as supported model providers are not covered there. The tool targets developers who want a CLI coding agent that adapts to their own conventions instead of requiring a hand-written rules file.
https://github.com/CommandCodeAI/command-code
DeepSeek Harness (dsh) is an open-source agent harness from DeepSeek AI built on an everything-is-a-plugin architecture over the Cordis framework, started from npm with a local web UI.
Open-source terminal coding agent written in Rust that edits files, runs commands and works toward a goal across many model providers.
Tabby is a self-hosted, open-source AI coding assistant providing a private and customizable alternative to GitHub Copilot with rich integration and deployment options.
Cursor is an AI-powered code editor designed to enhance programming productivity by integrating intelligent code suggestions and AI-driven features directly into the coding workflow.
Serena is a free and open-source coding agent toolkit that enhances large language models with semantic code retrieval and editing capabilities, enabling autonomous coding workflows directly on your codebase.
Open-source terminal coding agent from the Qwen team with subagents, skills, hooks and MCP support, speaking the OpenAI, Anthropic, Gemini and Qwen API formats plus local model runtimes.
Plandex is an open-source AI coding agent designed to plan, execute, and manage large-scale coding tasks and real-world projects with high efficiency and flexibility.
Smol Developer is a library and toolset that embeds a junior developer AI agent into applications to scaffold and generate codebases interactively through human-in-the-loop prompt engineering.