What does this skill do, and when should you use it?
This skill guides agents in using the Sysinternals DebugView CLI, DbgViewCli, to capture Win32 OutputDebugString and kernel DbgPrint/KdPrint output on Windows. It supports filtering by wildcard pattern, PID, or process name and can emit text, CSV, or XML. The workflow also covers bounded capture, boot logging, remote monitoring, log files, and crash-dump analysis for debug output. Kernel and boot features require administrator privileges, and the skill is Windows-only.
It instructs an agent to locate and verify dbgviewcli.exe, check instance status, start Win32 or kernel capture, filter output by pattern, PID, or process name, and write results to stdout or a log file. It also explains bounded execution with duration, line-count, match-pattern, and tail limits; boot logging; remote connections; pause, resume, and stop controls; and analysis of relevant crash dumps.
- A Windows application developer needs OutputDebugString capture without relying on an application logging framework.
- A driver developer needs administrator-level collection of DbgPrint or KdPrint kernel output.
- An automated debugging workflow must stop after a fixed duration, line count, or matching error appears.
- A troubleshooter needs to isolate output from one PID, process name, or wildcard pattern.
- An operator needs boot-time kernel logging or remote monitoring of a DbgView instance.
How do you install this skill?
- Remote monitoring uses unauthenticated, unencrypted TCP and scans ports 2020–2030; avoid untrusted networks.
- Kernel capture and boot logging load a driver or change persistent boot configuration; confirm administrator scope and recovery steps before use.
- The binary, scripts, driver compatibility, and parameter behavior were not executed or independently validated in this static review.
- Captured output may contain sensitive debugging information, but the skill defines no redaction, access-control, or retention policy.
- Shell / CLI
- Network access
- Local filesystem
Windows Vista or laterdbgviewcli.exeAdministrator privileges for kernel or boot capture
Place dbgviewcli.exe on PATH or invoke it by full path. Before execution, verify the file with PowerShell's Get-AuthenticodeSignature and require an Authenticode signature from Microsoft Corporation; reject unsigned files or files signed by another publisher. The repository collection can be installed with npx skills add microsoft/skills, then selecting debugview in the wizard. The README states that skills are installed in the selected agent directory, such as .github/skills/.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/microsoft/skills.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/.github/skills/debugview" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Start with dbgviewcli --status. For bounded Win32 capture, run dbgviewcli --no-banner --duration 30. To capture until an error appears, run dbgviewcli --no-banner --wait-for "*ERROR*" --max-lines 10000. Kernel capture requires administrator privileges; an example is dbgviewcli --kernel --no-banner --duration 60 --format csv --log kernel_debug.csv. Example agent prompt: On Windows, use DebugView to capture debug output from myapp.exe for 30 seconds, keep the last 50 lines, and output CSV.
What are this skill's strengths and limitations?
- Designed for command-line, scripted, and agent-driven debug capture.
- Covers Win32 and kernel debug output, filtering, logging, remote monitoring, and boot logging.
- Provides duration, max-lines, wait-for, and no-banner controls suited to automation.
- Supports text, CSV, and XML output plus machine-readable status checks.
- Windows Vista or later is required; non-Windows platforms are excluded.
- Kernel capture and boot logging require administrator privileges and the Dbgv.sys driver.
- It is a standalone native executable with no package-manager installation path; the binary must be obtained and signature-verified separately.
- It is not intended for log4j, Serilog, Azure Monitor, ETW tracing, or user-mode crash dumps.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| DebugView CLI Capture this page ✓ Microsoft · Official | 50 · Use with care | ★ 2.7k | 3mo ago | MIT |
| VS Code Crash Dump Symbolicator ✓ Microsoft · Official | 53 · Use with care | ★ 194k | 11d ago | MIT |
| Browser DevTools Testing | 59 · Recommended | ★ 107k | 12d ago | MIT |
| Agent Host Log Investigator ✓ Microsoft · Official | 55 · Use with care | ★ 194k | 11d ago | MIT |
| ce-debug: Systematic Bug Diagnosis & Fix | 55 · Use with care | ★ 26k | 11d ago | MIT |
For ETW tracing, the skill directs users to WPR/xperf; for user-mode crash dumps, it directs users to WinDbg. It also explicitly excludes application logging frameworks and Azure Monitor or cloud telemetry.
How did FollowSkills review this skill?
The documentation requires Microsoft Authenticode verification, distinguishes standard-user and administrator operations, recommends bounded capture, and discloses that remote TCP monitoring is unencrypted and unauthenticated. However, remote connections, driver loading, boot-registry changes, and EULA registry writes lack explicit per-action user confirmation, sensitive-data handling guidance, and a complete rollback procedure, so points are deducted.
Parameters, workflows, output formats, execution bounds, and troubleshooting are reasonably documented. However, the claimed binary and scripts are not provided, and no debugview-specific tests establish behavior, malformed-input handling, driver compatibility, or diagnostic error reporting; the score is therefore limited by static calibration and reduced for uncertainty.
The audience, trigger phrases, Windows requirements, privilege boundaries, and non-fit scenarios are clearly stated. The skill is Windows-only, provides no Chinese-language or localization guidance, and remote use depends on a target-side DbgView instance and fixed ports, leaving some boundary and environment-fit uncertainty.
The skill has installation notes, requirements, progressive parameter documentation, examples, architecture, references, best practices, and troubleshooting. The repository supplies MIT licensing and verified official-organization provenance, but the skill lacks its own version, changelog, named maintainer, or explicit update path; listed bundled resources are not included in the supplied evidence, so points are deducted.
The skill directly addresses Win32/kernel debug capture, filtering, formatting, logging, bounded execution, and remote monitoring, with output schemas that appear directly consumable. Static evidence contains no execution results for the skill, and users must separately provide the binary, driver, and suitable privileges, so the score remains near but below the static ceiling.
The SKILL.md and three reference files provide auditable parameter, protocol, structure, and output-format details, while the repository includes generic CI and harness evidence. There are no skill-specific committed tests, real run outputs, or independent corroboration, preventing a higher static score.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →