Learn16 guides
How agentic coding works.
The systems behind context selection, tools, memory and safe edits.
01What Is Context in an AI Coding Assistant?Define the information an AI coding assistant can actually use for a task.02Context Window vs Codebase ContextWhy a large token window does not equal understanding a large repository.03Why AI Coding Agents Forget Your CodeUnderstand context truncation, changing files and conversation summaries.04RAG vs Full Context for Coding AgentsCompare retrieval with dumping an entire repository into one prompt.05Why More Context Can Make Coding Agents WorseUnderstand noise, contradictory evidence and token cost in coding agents.06How Licuse Selects ContextA transparent view of Licuse's task context, local history and remote AI path.07Building an Episode-Based Context System for Coding AgentsDesign notes on task-sized memory for a coding agent.08What Is an Agentic IDE?Define an IDE where AI can inspect, edit and verify a real project.09AI Code Editor vs AI Coding AgentSeparate the workspace interface from the actor that edits code.10AI IDE vs CLI Coding AgentCompare visual and terminal surfaces for agentic development.11What Is MCP in AI Coding?Understand how Model Context Protocol connects agents to tools and data.12What Is ACP?Learn how Agent Client Protocol connects coding agents to editors.13How Coding Agents Edit Multiple FilesA safe workflow for agent changes that span a repository.14How AI Agents Read Large CodebasesUnderstand repository maps, search and selective file reading.15How Coding Agents Run Terminal Commands SafelyPractical permission boundaries for agent shell access.16What Is Agent Planning in AI Coding?Understand how agents break a coding task into verifiable steps.
Explore another subject