Hacker News 熱門

Show HN: I built a post-mortem debugger for native Windows x64/x86 crashes

2026/09/24 03:15· 20天前

AI 摘要

Hello HN! I've spent years debugging Windows crashes with tools that were either friendly but limited (e.g. Visual Studio) or powerful but archaic (e.g. WinDbg). I developed patterns and methods for understanding what was going on, and decided to build it into a much more effective debugging tool called ForensicDbg. I built a modern interface to minimize the friction when debugging. All of the data shown to you is analyzed, interpreted, and presented to you clearly, so you can focus on what matters. Everything is interlinked so you can quickly and intuitivly navigate through the process space. ForensicDbg comes with an MCP server which allows for agenic debugging. The work done to interpret and interlink your data also benefits AI tools. It removes the risk of hallucinations while building a stable foundation for them to work from without spending tokens. If you want to try it out you can sign up and get a free beta license here: https://www.forensicdbg.com/beta Comments URL: https://news.ycombinator.com/item?id=49821086 Points: 3 # Comments: 0

正文

Visual Debugging

AI Debugging

ForensicDbg is a modern post-mortem debugger, designed from the ground-up to let you quickly and easily track down and identify your hardest to catch bugs.

We are currently running a private Beta test. If you are interested in debugging and want to ensure ForensicDbg surpasses your needs, request an invite.

Dark

Advanced user-mode post-mortem debugger for Windows

Debug x86 and x64 crash dumps
Attach to live processes
Set as just-in-time debugger to debug crashes when they happen
Simple C++ expression evaluator
SourceServer and SourceLink support
In active development with regular updates and feature additions

MCP interface for automated crash analysis with AI tools

Interoperates with all AI tools that support stdio MCP Server communication
Output is tailored for efficient consumption by AI tools
Higher quality data means better results from the AI
Interprets and labels data so the LLM doesn’t spend time doing it
Reduces overall token costs while unlocking better results from your AI tools

State-of-the-art UI built to be intuitive and responsive

 Color-coded output is understandable at a glance
Navigate the process space by clicking on any address
Memory regions are labeled with symbol names and object types
Objects displayed in structure initialization format
Multi-threaded symbol loading
Panels are kept in-sync as you navigate
Light and dark modes that match your system

Data is interpreted using decades of expert experience

Deduces types based on usage and references
Tracks register values as they flow through disassembly
Validates callstacks to remove misleading references
Simulates image loader to reconstruct missing read-only regions in mini-dumps
Automatically selects the thread, frame, and instruction that caused the exception

ForensicDbg gives AI tools access to its powerful debugging engine through its MCP server, greatly improving the quality of results and significantly lowering the token cost of tracking down your hardest to catch bugs.

We are currently running a private Beta test. If you are interested in debugging and want to ensure ForensicDbg surpasses your needs, request an invite.

Bug

CTextWindowBase::SetSelectionBegin doesn't null-check the CRefPtr<CTextLine> returned by GetLine() before dereferencing it to fetch m_link. This is only reachable when the window has no lines (GetLineCount() == 0) at the moment of a click — here, a disassembly panel that hasn't yet had a function/frame's code loaded into it (no region, no symbol walk, no frame selected). The fix is to guard the GetLine() result for null in SetSelectionBegin (CTextWindow.cpp around line 559) and skip the selection-link update when the window is empty.

Advanced user-mode post-mortem debugger for Windows

Debug x86 and x64 crash dumps
Attach to live processes
Set as just-in-time debugger to debug crashes when they happen
Simple C++ expression evaluator
SourceServer and SourceLink support
In active development with regular updates and feature additions

MCP interface for automated crash analysis with AI tools

Interoperates with all AI tools that support stdio MCP Server communication
Output is tailored for efficient consumption by AI tools
Higher quality data means better results from the AI
Interprets and labels data so the LLM doesn’t spend time doing it
Reduces overall token costs while unlocking better results from your AI tools

State-of-the-art UI built to be intuitive and responsive

 Color-coded output is understandable at a glance
Navigate the process space by clicking on any address
Memory regions are labeled with symbol names and object types
Objects displayed in structure initialization format
Multi-threaded symbol loading
Panels are kept in-sync as you navigate
Light and dark modes that match your system

Data is interpreted using decades of expert experience

Deduces types based on usage and references
Tracks register values as they flow through disassembly
Validates callstacks to remove misleading references
Simulates image loader to reconstruct missing read-only regions in mini-dumps
Automatically selects the thread, frame, and instruction that caused the exception

查看原文（Hacker News 熱門）