References: AI Coding Harnesses and Agentic Loops

1. Software agent - Wikipedia - Foundational coverage of autonomous software agents including the perceive-decide-act loop that AI coding harnesses extend with LLM tool use.

2. Integrated development environment - Wikipedia - Background on the developer-tool category that AI coding harnesses are evolving from; useful for understanding why harnesses bind file-system and shell tools to the model.

3. Tool use in artificial intelligence - Wikipedia - The broader AI article covers tool-using agents and the function-calling pattern central to this chapter's harness discussion.

4. AI Engineering - Chip Huyen - O'Reilly - Chapters on agents, tool use, and multi-step reasoning provide the vendor-neutral framing for what Claude Code, Codex, and Antigravity each do differently.

5. Building Effective Agents - Anthropic Engineering Team - Anthropic Blog (December 2024) - Authoritative essay on agent vs workflow patterns that defines the vocabulary used in this chapter for tool loops, subagents, and orchestration.

6. Claude Code Documentation - Anthropic - Authoritative reference for Claude Code installation, hooks, MCP servers, and session management covered in this chapter and Chapter 18.

7. OpenAI Codex CLI - OpenAI GitHub - The Codex CLI repository with installation, configuration, and usage; the OpenAI counterpart to Claude Code.

8. Google Antigravity - Google - Official site for the Antigravity agentic IDE; reference for the Google entrant in the coding-harness space discussed in this chapter.

9. ReAct: Synergizing Reasoning and Acting - Yao et al. (arXiv) - The 2022 paper that defined the reason-act-observe pattern used by every modern coding harness; short and worth reading once for primary-source grounding.

10. Anthropic Engineering Blog: How we built Claude Code - Anthropic Engineering - Posts on context management, tool design, and agent loop control that go behind the scenes of the harness discussed in this chapter.