hackhaton-space-stackHackhaton Space Stack
AcademyBuilder
Get Started
All terms
The Basics
  • Stack
  • Frontend & Backend
  • CLI
  • Monorepo
  • Server
  • Localhost
How AI Works
  • Context Window
  • Hallucination
  • Token
  • Prompt Caching
  • Session
  • Compaction
  • Embedding
  • Vector Database
  • RAG
  • Fine-tuning
  • Temperature
  • Inference
  • Reasoning
  • Multimodal
Building With AI
  • Agent
  • MCP
  • System Prompt
  • Skill
  • CLAUDE.md
  • Slash Command
  • Harness
  • Computer Use
  • Agents SDK
  • Voice Agents
  • OAuth
  • Vibe Coding
  • Permission Scope
  • Tool Calling
  • Prompt Injection
  • Eval
  • Guardrails
  • Sandbox
  • Progressive Disclosure
Code & Collaboration
  • Git
  • Commit
  • Branch
  • GitHub
  • Pull Request
  • Open Source
  • Markdown
  • Dependency
  • Merge
  • Fork
APIs & Connections
  • API
  • Auth
  • Database
  • ORM
  • SDK
  • Webhook
  • Endpoint
  • REST
  • HTTP Methods
  • Env File
  • Schema
  • JSON
  • YAML
  • Secret
  • Rate Limit
  • CORS
  • Cookie
  • Encryption
Shipping & Running
  • Deploy
  • Headless
  • Cron
  • DNS
  • CDN
  • Object Storage
  • Serverless
  • Edge
  • Worker
  • Runtime
  • Process
  • Daemon
  • Queue
  • Job
  • State
  • Cache
  • SSH
  • Build
  • Staging
  • Rollback
  • Docker
  • Feature Flag
  • Test
  • CI/CD
  • The Cloud
Debugging & Errors
  • Trace
  • Type Error
  • Stack Trace
  • Log
  • Bug
  • Patch
  • Latency
How Developers Think
  • DRY
  • YAGNI
  • KISS
  • Refactoring
  • Technical Debt
  • Async
← All terms

Type-safe, modern TypeScript scaffolding for full-stack web development

ThreadsGitHub

Info

  • Academy
  • Docs

Legal

  • Terms of Service
  • Privacy Policy

© 2026 Dzulhelmy Nazri

Building With AI

$defineharness--plain-english

Harness

TLDRThe car built around the AI engine.

The model is a brilliant intern with no badge, no laptop, and no second step. The harness is the office you put them in.

Cursor, Claude Code, Codex — those are offices. They decide which tools exist, what the sandbox allows, how MCP plugs in, when to ask you before rm. The same brain in a different office behaves like a different employee.

Raw, the model predicts text. That is the whole trick. It cannot open a file, save a file, run a CLI, or take a second step unless something around it feeds the result back and says "now what?" That loop is the quiet magic. One reply and a full stop is a chatbot. A loop with hands is an agent.

Hands and a loop

Tool calling is the intern filling a work order. The harness is what actually runs it. npx create-hackhaton-space-stack@latest mcp is a pair of hands you bolt onto whichever office you already use. The plugin is the same hands plus skills and slash commands, installed into Claude Code or Codex.

Two products can run the identical model and feel like different tools. One office has memory, the right tools, a loop that finishes the job. The other handed you a brain in a jar and a wrench.

When someone says "the model can't do that," check the office first. Often the intern could. The harness never offered a desk.

What this unlocks

Great engines are cheap. The office is what turns one into something that actually works. Blame the intern for a hallucination. Blame the office when there was never a tool named scaffold_project.

A brain in a jar talks. An office ships a folder.

Related

  • Agent
  • Tool Calling
  • Sandbox
  • MCP
PrevSlash Command

Building With AI

NextComputer Use