mcp-windbg: Instructions file for Claude Code

CLAUDE.md

mcp-windbg CLAUDE.md is an instructions file for Claude Code from svnscha/mcp-windbg. It costs 3,289 tokens per session, scanned A, original, MIT.

An instruction file for mcp-windbg, a Windows-only server that lets AI models inspect crash dumps and control the WinDbg or CDB debuggers. It describes the required tools, transports, tests, and coverage checks.

In plain words
What is it for?
Use it when developing, testing, or running mcp-windbg, investigating Windows crashes, analyzing dump files, or connecting the server through standard input/output or HTTP.
Why use it?
It tells the coding agent how to run the project and which tests can work without a Windows debugger. It also explains that live debugging tests need CDB and that coverage must include both processes involved.

Instructions file for Claude Code

Written for Claude Code: Claude Code plugin machinery.

This is svnscha/mcp-windbg's own configuration. It tells Claude Code how to work on mcp-windbg itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything mcp-windbg configures →

About the project

mcp-windbg is a Model Context Protocol server that lets AI models control WinDbg, the Windows debugger used to inspect crash dumps and debug user-mode or kernel software. Developers and investigators use it to run debugger commands, examine failures, and manage multiple live debugging sessions through natural-language requests. The catalogue add-ons connect coding agents to this debugging workflow.

svnscha/mcp-windbg · 1,557 stars · on GitHub · svnscha.de

Reuse

Borrowing it

Nothing to install: this file belongs to svnscha/mcp-windbg. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/svnscha/mcp-windbg/develop/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/svnscha/mcp-windbg

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for mcp-windbg CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/svnscha/mcp-windbg/claude-md.svg)](https://agentmods.dev/instructions/svnscha/mcp-windbg/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/svnscha/mcp-windbg/claude-md"><img src="https://agentmods.dev/badge/instructions/svnscha/mcp-windbg/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 3,289 This file is loaded in full into every session.
When invoked 3,289 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.03289 $0.03289
Opus 5 $0.01644 $0.01644
Sonnet 5 $0.00658 $0.00658
Haiku 4.5 $0.00329 $0.00329

Measured 8d ago against content hash 344cba213753, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

mcp-windbg CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

CLAUDE.md · 228 lines

How it starts

The opening of the file, as written. The whole thing — 228 lines — stays where its author put it; the contents beside it link to each section on GitHub.

mcp-windbg

A Model Context Protocol server that lets AI models analyze Windows crash dumps and drive live or remote debugging through WinDbg/CDB. It is a Python wrapper around cdb.exe: the model calls MCP tools, the server runs the matching debugger commands and returns the text output. Windows-only (needs CDB). The single entry point is mcp-windbg, which speaks MCP over stdio or streamable-http.

Build / test / run

Python 3.10+ with the uv package manager. CDB must be installed for the live tests (WinDbg from the Microsoft Store, or the Windows SDK).

uv sync --dev                                                       # install incl. dev deps
uv run pytest src/mcp_windbg/tests/ -v                             # full test suite
uv run pytest src/mcp_windbg/tests/ -v -m "not live"              # hermetic subset (no CDB)
uv run python -m mcp_windbg --verbose                              # run the server (stdio)
uv run python -m mcp_windbg --transport streamable-http --port 8000   # HTTP transport

Coverage

The code under test runs in two processes, and both must be measured or the number lies:

  • the hosted server subprocess, where tool dispatch executes. Set MCP_WINDBG_COVERAGE so the harness launches it under coverage run --parallel-mode.
  • the pytest process itself, where the hermetic unit tests (tests/test_*.py) run. This needs coverage run -m pytest, not plain pytest. Miss this and every unit test reads as dead code: that is how kd_session.py once reported 34% while fully tested.

Both write parallel-mode data files that coverage combine merges:

uv run coverage erase
$env:MCP_WINDBG_COVERAGE = "1"
uv run coverage run -m pytest src/mcp_windbg/tests/     # measures pytest AND the server
$env:MCP_WINDBG_COVERAGE = $null
uv run coverage combine                 # merge the per-process .coverage.* files
uv run coverage report                  # or: uv run coverage html  ->  htmlcov/

Read the full file on GitHub · 228 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 8d ago First seen · 228 lines · 3,289 tokens per session scan A 344cba213753

Subscribe to this mod's changes

mcp-windbg CLAUDE.md is an instructions file published in the GitHub repository svnscha/mcp-windbg (1,557 stars, last pushed 2d ago), licensed MIT. It adds 3,289 tokens to every session, about $0.0164 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.