UtilKits
UtilKits.AI CODING LABS
BlogsKitsGlossary
About
UtilKits

Guides, updates, and insights on tools, technology, and useful resources.

Explore

  • All Articles
  • Kits
  • Glossary
  • Tools
  • LLM API Cost Calculator
  • guide
  • info
  • updates

Others

  • About
  • Contact
  • Privacy Policy
  • Terms of Service

© 2026 UtilKits

GitHub
  1. Home
  2. Glossary

Glossary

Plain-language definitions for agentic AI and coding-agent development. Reuse them in your own project.

Tags:Tags filter the list and glossary.md. With no tag selected, the file has every term.
Also applies to glossary.md. Turn it off to download the file without sources.
Export glossary.md

Showing 1–20 of 46 terms

glossary.md: 46 terms, with sources

Download glossary.mdThe file has every term matching the selected tags, not only the ones loaded on screen. The letter filter only changes what is on screen.

Agent teams

Several separate AI sessions that work together. One session acts as the team lead, and they share a task list and can message each other. Unlike a subagent, each teammate has its own context window.

Sources (2)
  • Claude Code docs: Run agent teams(opens in a new tab)
  • Autohand docs: Agent Teams(opens in a new tab)

Agentic coding

A way of working where the AI reads files, runs commands, and edits code by itself while you watch or step in. A chat assistant only replies with text, and you apply it yourself.

Sources (3)
  • Claude Code docs: How Claude Code works(opens in a new tab)
  • Scrimba: What Is Agentic Coding?(opens in a new tab)
  • Addy Osmani: AI Coding Agent(opens in a new tab)

Agentic harness

Everything around the model that turns it into a coding agent: file access, running commands, permission checks, memory loading, and the loop that chains steps together. The harness is the tool you run. The model is the AI inside it.

Sources (3)
  • Claude Code docs: How Claude Code works(opens in a new tab)
  • Endor Labs: What is an agent harness?(opens in a new tab)
  • UI Bakery: What Is an Agent Harness?(opens in a new tab)

Agentic loop

The cycle an agent repeats on every task: gather context, take an action, check the result, and go again until done. Each tool result guides the next step, and you can interrupt at any time. Hooks, skills, and MCP plug into different parts of this loop.

Sources (3)
  • Claude Code docs: How Claude Code works(opens in a new tab)
  • OpenAI Agents SDK: Running agents(opens in a new tab)
  • Atlan: What is an agent loop?(opens in a new tab)

AGENTS.md

A markdown file where you write instructions for AI coding agents. If a project has one and no CLAUDE.md, Claude Code reads it as the project instructions.

Sources (3)
  • Claude Code docs: Memory(opens in a new tab)
  • AGENTS.md(opens in a new tab)
  • Addy Osmani: AGENTS.md(opens in a new tab)

Artifact

A live web page that the agent publishes from your session to a private link, so you can look at the result visually or share it instead of reading terminal text. The page updates in place when the session republishes it.

Sources (2)
  • Claude Code docs: Share session output as artifacts(opens in a new tab)
  • TechCrunch: Gemini Canvas(opens in a new tab)

Auto memory

Notes the agent writes for itself from your corrections and preferences. They are saved per repository and loaded at the start of each session. It is the agent-written counterpart of CLAUDE.md, which you write yourself.

Sources (2)
  • Claude Code docs: Memory(opens in a new tab)
  • VS Code docs: Memory in agents(opens in a new tab)

Auto mode

A permission mode where a separate classifier model reviews actions instead of you. Most actions run without asking, while risky ones, like going beyond the task or using untrusted infrastructure, are blocked.

Sources (2)
  • Claude Code docs: Permission modes(opens in a new tab)
  • OpenAI Codex docs: Agent approvals & security(opens in a new tab)

Bare mode

A start-up option that skips hooks, skills, plugins, MCP servers, memory, and CLAUDE.md. It gives the same result on every machine, which suits CI and scripts.

Sources (2)
  • Claude Code docs: Run Claude Code programmatically(opens in a new tab)
  • OpenAI Codex docs: Non-interactive mode (--ignore-user-config, --ignore-rules)(opens in a new tab)

Bundled skills

Ready-made skills that ship with the tool, such as /batch, /debug, and /loop. Instead of running fixed logic, they give the agent a detailed prompt and let it plan the work.

Sources (2)
  • Claude Code docs: Skills(opens in a new tab)
  • OpenAI Codex docs: Agent Skills(opens in a new tab)

Checkpoint

A restore point saved at each prompt you send. The tool snapshots files before every edit, so you can roll code, the conversation, or both back to an earlier point. It is separate from git and does not track changes made through shell commands.

Sources (2)
  • Claude Code docs: Checkpointing(opens in a new tab)
  • Cursor docs: Agent overview (checkpoints)(opens in a new tab)

.claude directory

The folder where the agent reads project settings: settings, hooks, skills, subagents, rules, and memory. A project keeps one at its root, and your personal defaults live in a folder in your home directory.

Sources (3)
  • Claude Code docs: The .claude directory(opens in a new tab)
  • OpenAI Codex docs: Config basics(opens in a new tab)
  • Cursor docs: Rules(opens in a new tab)

CLAUDE.md

A markdown file of standing instructions you write for the agent. It is loaded at the start of every session. Put project conventions, architecture notes, and "always do X" rules here. It survives compaction. The agent can also read AGENTS.md.

Sources (3)
  • Claude Code docs: Memory(opens in a new tab)
  • AGENTS.md(opens in a new tab)
  • Cursor docs: Rules(opens in a new tab)

Cloud session

A session that keeps running after you close your laptop because it runs on cloud servers instead of your machine. A session in your own terminal or editor is a local session.

Sources (2)
  • Claude Code docs: Use Claude Code in the cloud(opens in a new tab)
  • GitHub Docs: Copilot cloud agent(opens in a new tab)

Command

A reusable instruction you run by typing /name. Built-in commands such as /clear and /compact control the session, and you can add your own. Skills are the recommended way to package multi-step commands.

Sources (2)
  • Claude Code docs: Commands(opens in a new tab)
  • OpenAI Codex docs: Slash commands(opens in a new tab)

Compaction

Automatic summarizing of the conversation when the context window is nearly full. Old tool outputs are cleared first, then the conversation is summarized. Instructions that only exist in the chat may be lost, but CLAUDE.md is reloaded.

Sources (2)
  • Claude Code docs: Explore the context window(opens in a new tab)
  • PromptLayer: Context compaction(opens in a new tab)

Context window

The working memory of a session. It holds the conversation, file contents, command output, CLAUDE.md, loaded skills, and system instructions. As it fills up, compaction summarizes it.

Sources (2)
  • Claude Code docs: Explore the context window(opens in a new tab)
  • IBM: What is a context window?(opens in a new tab)

Effort level

A setting that controls how much the model thinks on each step. Higher effort means more thinking and deeper reasoning. Lower effort is faster and cheaper.

Sources (2)
  • Claude Code docs: Model configuration(opens in a new tab)
  • OpenAI docs: Reasoning models(opens in a new tab)

Extended thinking

Step-by-step reasoning the model does before it answers. You can tune it with the effort level. The thinking is shown in the terminal in gray text.

Sources (3)
  • Claude Code docs: Model configuration(opens in a new tab)
  • Google Cloud docs: Gemini thinking(opens in a new tab)
  • OpenAI docs: Reasoning models(opens in a new tab)

Hook

A handler you define that runs automatically at a fixed point in the agent lifecycle, such as before a tool runs, after a file edit, or at session start. A hook can run a shell command, call a web endpoint, or run another agent. Hooks are deterministic: they fire at set points, not at the model discretion.

Sources (2)
  • Claude Code docs: Get started with hooks(opens in a new tab)
  • Cursor docs: Hooks(opens in a new tab)