Entries tagged :: ai

Build an autonomous crash triage workflow with the Claude Agent SDK, mcp-windbg, and Markdown history.

A standalone, code signed mcp-windbg.exe for machines without Python, and a fix for remote sessions whose target is already running when you connect.

Failed opens no longer leak debugger sessions, a timed-out command on a dump gets cancelled instead of wedging the session, and completion markers are unique per session. Plus what I learned measuring six community pull requests against a real debugger.

A quick follow-up to 1.3.1: open kernel crash dumps with open_kd_dump, and load your debugger extensions automatically with init commands.

Break-in while a command runs, partial output on timeouts, and a debugger that exits is reported right away. Three of the four fixes came from the community.

The MCP server, Claude Code skills, and crash-analysis agent are now separate plugins for use with your preferred installation.

A fix for missing Unicode output on Chinese, Japanese, and Korean Windows systems.

mcp-windbg now ships as a Claude Code plugin with four debugging skills and a crash-analysis agent.

Support for the 2.x MCP SDK, a new wait_for_break tool, and fixes for 'g' freezing the machine it was supposed to release.

Kernel debugging, session IDs, and a redesigned tool surface. The first stable release.

Break into live sessions, redact sensitive output before it reaches the model, and a brand-new documentation site. A year and a month in, with 1.4k stars and a community that keeps showing up.

I built an autonomous AI agent that works from a PRD, verifies its own work, and ships code while I sleep. Here's Ralph.

Ollama v0.14 supports the Anthropic Messages API, allowing Claude Code to work with local models.

Exposing .NET agents through OpenAI-compatible endpoints and connecting them to LibreChat.

My first experiments with NVIDIA's DGX Spark, including local models with Ollama, LibreChat, and ComfyUI.

Taking our agent from 'hello world' to useful - with database persistence, embeddings, vector search, and tools that let it reason over documentation and conversation history.

Introducing the Knowledge repository - a hands-on companion for learning how to build AI agents with the Microsoft Agent Framework in C#.

Adding Prompts API support and streamable-http transport to mcp-windbg.

Here's what nobody tells you about AI-assisted rapid prototyping: Every prototype feels like success, but success is measured by shipping, not building.

AI-powered crash analysis gets a major upgrade with PyPI availability, live debugging sessions, and community-driven improvements. One-command installation is finally here.

Because copy-pasting prompts from your notes app is so last year. Let me show you how to turn your best prompts into reusable superpowers.

Because manually squinting at hex dumps is so last century. Let me show you how AI-assisted debugging is leaving WinDBG's command line in the dust.

Because watching GitHub Copilot refactor files with a single prompt is the most satisfying thing you'll see today.

Because running your own AI is the real power move.