agent-shell and ai-code-interface.el
About agent-shell
xenodium/agent-shell
A native Emacs buffer to interact with LLM agents powered by ACP
Communicates with ACP-compliant agents (Claude, Gemini, Mistral, etc.) through stdio transport via the acp.el library, enabling multi-turn conversations within Emacs buffers. Extensible through a rich ecosystem of companion packages for features like session management, notifications, code review interfaces, and multi-agent coordination. Built on comint-shell infrastructure for reliable terminal-like interaction with LLM agents.
About ai-code-interface.el
tninja/ai-code-interface.el
Unified Emacs interface supporting OpenAI Codex, GitHub Copilot CLI, Claude Code, Gemini CLI, Opencode, and more
It provides a uniform Emacs user experience for various AI coding CLIs and external backend packages, focusing on a consistent workflow rather than deep integration with a single agent. The project leverages Emacs's built-in MCP (Major Mode Context) tools, `imenu`, `xref`, and `tree-sitter` for context gathering, prompt construction, and refactoring actions. It targets iterative development, offering features like a TDD cycle, diff review, Magit helpers, and multi-session support per backend or Git worktree.
Scores updated daily from GitHub, PyPI, and npm data. How scores work