MCP server · Audited from source
mcp-windbg
by svnscha/mcp-windbgMIT licenceAudited 24 Oct 2026 · source 6c3c8fd69867
Is it safe? Yes — SAFE, grade B (89/100). It exposes 12 tools: 9 read, 3 write and 0 destructive. The audit recorded 8 findings (0 critical, 0 high, 0 medium, 6 low). Maintainer's description: Model Context Protocol for WinDbg.
Cite
Cite this page
ModelsAtlas. “mcp-windbg (svnscha/mcp-windbg)”. Retrieved 24 Oct 2026. https://modelsatlas.com/mcp/svnscha/mcp-windbg
Trust audit
Audit 0.4.1 · full source read 24 Oct 2026
8 findings, listed below with the file and line each was found on. Grade A needs a sandboxed behavioural run too; this audit read the source only.
What the source does
- FilesystemDeclared2 observations
- NetworkDeclared2 observations
- ShellNone observedNothing seen in the source
- DependenciesPinnedNothing floats
- SecretsNone foundNo keys in source
“Declared” means the server says it does this and the source agrees. Browser, file and shell access are expected of some kinds of server.
Five audit layers
| Layer | What we check | Result |
|---|---|---|
L0 Provenance | Repo owner, licence, archive status, activity | PASS 3 findings |
L1 Static source | Secrets, path traversal, supply chain | PASS 5 findings |
L2 Instructions | Prompt injection and authority framing in docs and tool descriptions | PASS |
L3 Inventory | Declared tools match the source | PASS |
L4 Sandbox | Run in isolation and watch behaviour | SKIPPED |
Provenance & runtime
Who published it, and what the audit could and could not run
Provenance
- Repository owner
svnschaUser accountAccount type recorded when we read the repository. - PackagePublished on PyPI
mcp-windbgat version 1.6.0 - LicenceMITFrom the repository, as read at the audited commit.
- Last pushTodayActivity of the repository when we read it.
- Source commit read6c3c8fd69867The audit covers exactly this commit, read on 24 Oct 2026.
Runtime
- Sandbox runNot run yetThis audit read the source only; it did not execute the server.
- Tool inventory12 declaredRead from the source and classed read, write or destructive.
Every value here is read from the record's own repository and audit.Suggest a correction
Findings (8)
Sorted by severity
LOWNetwork egressnet.raw_ip · CLAUDE.md:285
python -m mkdocs serve # live preview at http://127.0.0.1:8000
- Why
- No further detail recorded for this rule.
- Fix
- —
LOWNetwork egressnet.raw_ip · docs/reference/cli.md:58
The endpoint is then `http://127.0.0.1:8000/mcp`. See
- Why
- No further detail recorded for this rule.
- Fix
- —
LOWNetwork egressnet.raw_ip · docs/scenarios/http-service.md:13
It serves MCP at `http://127.0.0.1:8000/mcp`. Pass the same server options as usual, for example
- Why
- No further detail recorded for this rule.
- Fix
- —
LOWNetwork egressnet.raw_ip · requirements-docs.txt:3
# python -m mkdocs serve # live preview at http://127.0.0.1:8000
- Why
- No further detail recorded for this rule.
- Fix
- —
LOWNetwork egressnet.raw_ip · scripts/test_signing_http.py:55
SIGNING_URL=f"http://127.0.0.1:{server.server_port}/sign")- Why
- No further detail recorded for this rule.
- Fix
- —
INFOInventory / provenanceinv.oversize · src/mcp_windbg/tests/dumps/DemoCrash1.exe.7088.dmp
src/mcp_windbg/tests/dumps/DemoCrash1.exe.7088.dmp
- Why
- 8190443 bytes not read
- Fix
- —
INFOInventory / provenanceinv.oversize · src/mcp_windbg/tests/dumps/KernelMiniDump.0xE2.dmp
src/mcp_windbg/tests/dumps/KernelMiniDump.0xE2.dmp
- Why
- 1562596 bytes not read
- Fix
- —
Tools it exposes
12 declared in source · 0 destructive
| Tool | Risk | What it does |
|---|---|---|
run_cdb_command | write | No description in the source |
run_kd_command | write | No description in the source |
send_ctrl_break | write | No description in the source |
close_cdb_session | read | No description in the source |
close_kd_session | read | No description in the source |
list_dumps | read | No description in the source |
open_cdb_dump | read | No description in the source |
open_cdb_remote | read | No description in the source |
open_kd_dump | read | No description in the source |
open_kd_session | read | No description in the source |
read_debugger_log | read | No description in the source |
wait_for_break | read | No description in the source |
Install
Package version 1.6.0 · audited commit 6c3c8fd
claude mcp add mcp-windbg -- uvx mcp-windbg==1.6.0
claude mcp add mcp-windbg -- uvx mcp-windbg==1.6.0
The command pins the package version recorded in the source (1.6.0); the audit covers commit 6c3c8fd69867. Transport: stdio · streamable-http. No environment variables required.
Audit history
- ✓SAFE B 89 / 100
First audit ·
6c3c8fd69867
README badge
For maintainers · updates on every re-audit
[](https://modelsatlas.com/mcp/svnscha/mcp-windbg/)
The badge always shows the latest audit; a maintainer cannot pin an old verdict.
Other ratings
Cited, never blended into our score
No third-party security ratings are recorded for this server yet. When they exist they appear here with their source and date, beside, not inside, our grade.
Questions about mcp-windbg
What is the mcp-windbg MCP server?
What tools does mcp-windbg expose?
Is mcp-windbg safe to connect to an agent?
What credentials does mcp-windbg need?
How does mcp-windbg run?
How current is this page?
Sources
- svnscha/mcp-windbg on GitHub — the source we read, its licence, stars and activityObserved · 24 Oct 2026
- ModelsAtlas trust audit — SAFE verdict for commit 6c3c8fd69867Observed · 24 Oct 2026
Spotted something wrong? Report a correction · Methodology
Keep going in your AI
Opens with this page as the source. We add nothing else to the prompt.
- What is the mcp-windbg MCP server?
- What tools does mcp-windbg expose?
- Is mcp-windbg safe to connect to an agent?