/code

Prerequisites

Emacs with libxml2. Check: M-: (libxml-available-p) should return t.

Step 1: Install

;; With straight.el
(use-package codeium
  :straight (:host github :repo "Exafunction/codeium.el")
  :init
  (add-to-list 'completion-at-point-functions #'codeium-completion-at-point)
  :config
  (setq use-dialog-box nil))

;; With quelpa
(use-package codeium
  :quelpa (codeium :fetcher github :repo "Exafunction/codeium.el")
  :init
  (add-to-list 'completion-at-point-functions #'codeium-completion-at-point))

Step 2: Install Language Server

M-x codeium-install

Step 3: Authenticate

M-x codeium-init
;; Browser opens -> sign in with Google/GitHub/email
;; Token auto-applies, or paste manually

Verify

M-x codeium-diagnose

# Debian/Ubuntu
sudo apt install libxml2-dev
# Then reinstall Emacs or use emacs-gtk package

Prerequisites

• Emacs 27+
• Node.js 22+ (critical - older versions fail silently)
• GitHub Copilot subscription or free tier

Step 1: Install

;; Emacs 29+ with use-package + vc
(use-package copilot
  :vc (:url "https://github.com/copilot-emacs/copilot.el"
       :rev :newest :branch "main")
  :hook (prog-mode . copilot-mode)
  :bind (:map copilot-completion-map
         ("<tab>" . copilot-accept-completion)
         ("TAB" . copilot-accept-completion)
         ("M-n" . copilot-next-completion)
         ("M-p" . copilot-previous-completion)))

;; With straight.el
(use-package copilot
  :straight (:host github :repo "copilot-emacs/copilot.el")
  :hook (prog-mode . copilot-mode)
  :bind (:map copilot-completion-map
         ("<tab>" . copilot-accept-completion)))

Step 2: Install Server

M-x copilot-install-server

Step 3: Login

M-x copilot-login
;; Opens browser for GitHub OAuth

Step 4: Verify

M-x copilot-diagnose
;; Should show "Enabled" and your subscription status

# Check version (must be 22+)
node --version

# If using nvm:
nvm install 22 && nvm use 22

# If using system node (Debian/Ubuntu):
curl -fsSL https://deb.nodesource.com/setup_22.x | sudo -E bash -
sudo apt install -y nodejs

Prerequisites

• Emacs 29+ with native JSON
• plz 0.9+ and dash packages
• API key for chosen provider

Step 1: Install

;; From GNU ELPA (simplest)
(use-package minuet
  :ensure t
  :bind (("M-y" . minuet-complete-with-minibuffer)
         ("M-i" . minuet-show-suggestion))
  :hook (prog-mode . minuet-auto-suggestion-mode))

Step 2: Get API Key

Step 3: Configure Provider

;; Gemini (free, fast)
(setq minuet-provider 'gemini)
(plist-put minuet-gemini-options :api-key "GEMINI_API_KEY")
(plist-put minuet-gemini-options :model "gemini-2.0-flash")

;; Codestral (free, code-optimized)
(setq minuet-provider 'codestral)
(plist-put minuet-codestral-options :api-key "CODESTRAL_API_KEY")

;; Ollama (local, private)
(setq minuet-provider 'openai-fim-compatible)
(plist-put minuet-openai-fim-compatible-options
           :end-point "http://localhost:11434/v1/completions")
(plist-put minuet-openai-fim-compatible-options :api-key "TERM")
(plist-put minuet-openai-fim-compatible-options :model "qwen2.5-coder:3b")

Simple, extensible LLM client. Works anywhere in Emacs - any buffer, shell, minibuffer. Supports OpenAI, Anthropic, Gemini, xAI, Ollama, and OpenAI-compatible backends. Tool-use, MCP integration, multi-modal input (images, documents). v0.9.9 1100+ stars

(use-package gptel
  :config
  (setq gptel-model 'claude-sonnet-4-20250514
        gptel-backend (gptel-make-anthropic "Claude"
                        :key 'gptel-api-key-from-auth-source)))

Tool for interacting with LLMs. Translation, code review, summarization, grammar fixes. Built on the llm package. Great for local models via Ollama. v1.8.6 GNU ELPA 880+ stars

(use-package ellama
  :init
  (setopt ellama-provider
          (make-llm-ollama
           :chat-model "qwen2.5-coder:7b")))

Low-level library abstracting LLM capabilities. Used by ellama and others. Supports multiple providers with unified API. If building your own AI tools, start here.

Emacs-native AI assistant that understands and interacts with your codebase. From the creators of EAF and lsp-bridge. Spiritual successor to aidermacs. Uses tools to interact with environment based on LLM reasoning. Connects to various providers via LiteLLM. early stage

Integrates Aider into Emacs. Cursor-like experience. Native multiline prompts, Tramp support (SSH, Docker), session scratchpads. Works with Claude, DeepSeek, ChatGPT, local models. NonGNU ELPA

(use-package aidermacs
  :config
  (setq aidermacs-backend 'vterm)  ; or 'comint
  (global-set-key (kbd "C-c a") 'aidermacs-transient-menu))

Minimal Emacs UI for Aider. AI-driven agile workflows (TDD, refactoring). Bootstrapping utilities for new files and projects. v0.13.1

Anthropic's official CLI agent. Works in any terminal, integrates with any editor via file system. Powerful tool-use, web search, multi-file edits. Run via claude command.

Configuration system packaging domain knowledge into reusable components. 20 skills, 7 agents, 4 commands, 5 hooks. Skills auto-activate based on file context - open a .rs file and Rust patterns inject automatically.

Skills (20)

Agents (7)

Key Commands

# Install
cd kronael/assistants/claude-template
claude "install"
# Compares with ~/.claude/ and prompts before overwriting

Processes design specs through a multi-stage pipeline. Parallel task execution with configurable workers. Injects skills from ~/.claude/skills/ into worker prompts. Resumable with ship -c.

# Run ship on a design spec
ship design.md

# Resume interrupted run
ship -c

# Verbose mode
ship -v design.md

Docker container that runs takopi with Vite dev server for web deployments. Auto-discovers projects from /workspace. Nginx proxy to krons.fiu.wtf.

Any app_name/index.html in the web directory becomes available at https://krons.fiu.wtf/app_name/. No restart needed - Vite hot reloads.