---
title: "Introducing Debug Mode: Agents with runtime logs"
description: "Debug Mode is an agent loop for coding bugs that combines runtime information with human verification, addressing issues that can stump conventional coding agents. After the user describes a bug, it reads the codebase, proposes multiple hypotheses, and instruments the code with logging statements designed to test them. The user reproduces the issue while the agent collects variable states, execution paths, and timing information, then uses those logs to identify a root cause and produce a targeted fix, often a two- or three-line modification. The user reproduces the bug again with the fix; if it works, the agent removes its instrumentation, while a failed verification prompts additional logging and refinement until the user considers the problem solved."
---

# Introducing Debug Mode: Agents with runtime logs

[Cursor](https://yomu.fyi/company/cursor) · Alexey Kozy & Albert Slepak · Dec 10, 2025

**Type:** Announcement

## Summary

Debug Mode is an agent loop for coding bugs that combines runtime information with human verification, addressing issues that can stump conventional coding agents. After the user describes a bug, it reads the codebase, proposes multiple hypotheses, and instruments the code with logging statements designed to test them. The user reproduces the issue while the agent collects variable states, execution paths, and timing information, then uses those logs to identify a root cause and produce a targeted fix, often a two- or three-line modification. The user reproduces the bug again with the fix; if it works, the agent removes its instrumentation, while a failed verification prompts additional logging and refinement until the user considers the problem solved.

## Context

Some bugs consistently stump coding agents or require significant developer time because the agent lacks concrete runtime information and cannot independently judge whether a technically working fix feels right.

## Approach / What changed

Debug Mode reads the codebase, generates multiple hypotheses, instruments code with logging, collects runtime data while the user reproduces the bug, proposes a targeted fix, and uses another user reproduction to verify the result. It removes instrumentation after confirmation or adds logging and refines the approach when verification fails.

## Takeaways

- The agent generates multiple debugging hypotheses before attempting a fix, including possibilities the developer may not have considered.
- Runtime logs provide variable states, execution paths, and timing information that the agent uses to pinpoint the root cause and produce a targeted change.
- Human verification determines whether the fix is acceptable; failed verification triggers more instrumentation and another refinement cycle.

**Tags:** [AI Agents](https://yomu.fyi/topic/ai-agents), [Debugging](https://yomu.fyi/topic/debugging), [Human-in-the-Loop](https://yomu.fyi/topic/human-in-the-loop)

- Source: [Cursor](https://cursor.com/blog/debug-mode)
- Source URL: https://cursor.com/blog/debug-mode
- Ingested by Yomu: 2026-09-02T11:45:05.640Z

[Read original post](https://cursor.com/blog/debug-mode)