AI Agents
Abdulrhman-io IDE uses a team of specialized AI agents that collaborate to solve complex software engineering tasks.
Architecture
The Coordinator routes tasks to the right agents. Each agent has focused context, tools, and memory. They can hand off work, review each other, and iterate until the goal is complete.
Planner Agent
Analyzes the user request + project context, produces a step-by-step execution plan, identifies affected files, and estimates complexity.
Builder Agent
Implements the plan. Writes new code, modifies existing files, follows your style from Engineering Memory.
Debugger Agent
Analyzes errors, stack traces, and logs across 6 languages. Finds root cause and proposes or applies fixes.
Reviewer Agent
Reviews changes for correctness, security, performance, and style. Can block or suggest improvements before commit.
Security Agent
Scans for vulnerabilities, secret leaks, dangerous patterns. Enforces policies and can require human approval.
Coordinator
The orchestrator. Decides which agents to run, in what order, when to loop, and when to ask the user for approval.
You can customize agent behavior in Settings → AI Agents.