1. Coding Agents

1.1 Claude Code (Anthropic) — Agentic coding tool that lives in your terminal, IDE, or @claude on GitHub, editing files, running commands, and handling git workflows from natural language.

• Docs: https://docs.anthropic.com/en/docs/claude-code/overview
• Start here: https://www.anthropic.com/engineering/claude-code-best-practices

1.2 Codex (OpenAI) — OpenAI's coding agent spanning CLI, IDE, terminal, and a cloud/ChatGPT "command center" that runs agents in parallel across worktrees.

• Docs: https://openai.com/codex/ · Repo: https://github.com/openai/codex
• Start here: https://www.freecodecamp.org/news/the-codex-handbook-a-practical-guide-to-openai-s-coding-platform/

1.3 OpenCode — Open-source, terminal-first (MIT) coding agent that is bring-your-own-model (Anthropic, OpenAI, Gemini, local via Ollama) with Plan/Build modes.

• Docs: https://opencode.ai/docs/
• Start here: https://dev.to/rosgluk/opencode-quickstart-install-configure-and-use-the-terminal-ai-coding-agent-4kcb

1.4 OpenClaw — Viral open-source personal AI agent ("Molty") that runs locally, connects models to your apps/files, and can delegate to background coding agents (Claude Code, Codex, OpenCode).

• Docs/Repo: https://github.com/openclaw/openclaw
• Start here: https://www.digitalocean.com/resources/articles/what-is-openclaw

2. Claude Science (Anthropic)

Claude Science is Anthropic's local-first desktop "workbench" for rigorous research: a coordinating agent writes/runs Python/R/shell, pulls from 60+ scientific databases, and saves versioned artifacts with a full provenance trail.

• Docs: https://claude.com/docs/claude-science/overview
• Start here: https://techcrunch.com/2026/06/30/anthropics-claude-science-bets-on-workflow-not-a-new-model-to-win-over-scientists/

3. Scientific Agent Skills