The new paradigm for building powerful tools is to design them for AI models. Instead of complex GUIs, developers should create simple, well-documented command-line interfaces (CLIs). Agents can easily understand and chain these CLIs together, exponentially increasing their capabilities far more effectively than trying to navigate a human-centric UI.

The defining characteristic of an enterprise AI agent isn't its intelligence, but its specific, auditable permissions to perform tasks. This reframes the challenge from managing AI 'thinking' to governing AI 'actions' through trackable access controls, similar to how traditional APIs are managed and monitored.

Instead of placing agents inside a pre-set environment, a more powerful approach for reasoning models is to start with just the agent. Then, give it the tools and skills to boot its own development stack as needed, granting it more autonomy and control over its workspace.

Instead of creating a bespoke memory or messaging protocol for agent-to-agent communication, Notion leverages its core primitives. Agents compose by writing to and reading from shared Notion pages and databases, creating a decoupled, human-editable, and transparent system for coordination.

Traditional identity models like SAML and OAuth are insufficient for agents. Agent access must be hyper-ephemeral and contextual, granted dynamically based on a specific task. Instead of static roles, agents need temporary permissions to access specific resources only for the duration of an approved task.

Instead of designing tools for human usability, the creator built command-line interfaces (CLIs) that align with how AI models process information. This "agentic-driven" approach allows an AI to easily understand and scale its capabilities across numerous small, single-purpose programs on a user's machine.

The debate over whether CLI or MCP is better for AI agents is a false dichotomy. Both will coexist. The interface that attracts more token expenditure will likely improve faster due to reinforcement learning, but neither will fully displace the other.

"Code Mode" is not an alternative to MCP but a more efficient way to use it. Instead of multiple sequential tool calls, the model generates a single script that executes multiple actions in a sandbox. MCP still provides the core benefits of authentication, discoverability, and a standardized, LLM-friendly API.

Exposing a full API via the Model Context Protocol (MCP) overwhelms an LLM's context window and reasoning. This forces developers to abandon exposing their entire service and instead manually craft a few highly specific tools, limiting the AI's capabilities and defeating the "do anything" vision of agents.