← Read the full guide
All infographics

AI Agents & MCP

A chat model tells you what to do. An agent does it. MCP is how it does it.


5
Agent Components
100s
MCP Servers
4
Major Platforms
Real-World Tasks
The agent loop
while not done:
thought = model.think(history + tools)
// pick a tool, or finish
action = thought.next_tool_call
observation = tools.execute(action)
history.append(thought, action, observation)
The 5 components
REASONING
1. Model
Opus 4.7, GPT-5, Gemini Ultra. Smart enough to plan + self-correct.
2. Tools
Functions the agent can call. Each has a name, description, parameter schema.
3. Loop / harness
The orchestration code asking model what's next, executing, returning.
4. Memory
Working memory (this conversation) + optional long-term (RAG, vectors).
5. Safety / oversight
Confirmation gates, allow/deny lists, sandboxing, step budgets. Human-in-the-loop on destructive actions.
What is MCP?
Model Context Protocol — the open standard for connecting AI agents to external tools. USB for AI. Anthropic created it, every major vendor adopted it.
Top Failure Modes
Goal drift · Loop divergence · Hallucinated tool calls · Costly tangents · Prompt injection from external data
Pick by use case
High-value agent tasks
  1. Sustained coding
    → Claude Code
  2. Browser automation
    → ChatGPT Operator
  3. Google Workspace tasks
    → Gemini
  4. Real-time X / news
    → SuperGrok
  1. Multi-file refactors
    Across your codebase
  2. App Store metadata
    Build status, release notes
  3. Cross-tool workflows
    GitHub → Linear → Slack
  4. Continuous monitors
    Morning email summary