← Back to Skills Marketplace
ningtoba

Pc Assistant Fixed

by Ning · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
240
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install pc-assistant-fixed
Description
PC healthcheck and diagnostics with detailed system information and actionable recommendations. Works on Windows, macOS, and Linux. Read-only system diagnost...
Usage Guidance
This skill appears to be a legitimate Linux system healthcheck script, but before installing or scheduling it you should: (1) note the manifest and docs claim macOS/Windows support even though those platform scripts are not included — ask the author for the missing files or treat it as Linux-only; (2) review the healthcheck.sh yourself (it is included) and confirm you're comfortable with it reading files such as ~/.ssh/authorized_keys, ~/.ssh/known_hosts, /var/log/auth.log, and installed-package lists — these will appear in the generated reports; (3) run it as your regular (non-root) user and in a safe environment first (or inside a VM/container) to inspect outputs; (4) set PC_ASSISTANT_OUTPUT_DIR to a private directory with appropriate permissions and do not store reports in world-readable locations; (5) avoid scheduling until you confirm the script's behavior and the location/rotation/cleanup policy; and (6) if you expect true cross-platform support, request the missing macOS/Windows scripts or choose a different skill packaged for those OSes.
Capability Analysis
Type: OpenClaw Skill Name: pc-assistant-fixed Version: 1.0.0 The skill performs an overly broad system diagnostic that collects highly sensitive information, including the contents of SSH authorized keys, shell history (~/.bash_history), and all environment variables, which frequently contain API keys and secrets. While the script saves this data to local files in /tmp and does not contain explicit network exfiltration logic, the scope of data collection in scripts/healthcheck.sh is excessive for a general healthcheck and poses a significant security risk if the resulting reports are mishandled.
Capability Assessment
Purpose & Capability
The name/description promise cross-platform (Windows, macOS, Linux) diagnostics, but the provided repository only contains a Linux Bash healthcheck plus wrapper/scheduler. install.yaml and SKILL.md reference healthcheck.command (macOS) and healthcheck.ps1 (Windows) but those files are not present in the manifest — that is an inconsistency in claimed capability vs. delivered files.
Instruction Scope
The runtime instructions and scripts perform broad read-only system collection (hostname, /proc, lsblk, ip/ss, package lists, Docker, GPU, SMART, /var/log/auth.log, and snippets of ~/.ssh/authorized_keys and known_hosts). These actions are consistent with a diagnostics tool but will capture sensitive information (SSH key contents, failed-login lines, installed packages and user lists). The instructions do not attempt network exfiltration, but they do instruct writing timestamped reports to disk which could leak data if stored in world-readable locations.
Install Mechanism
There is no download-from-URL or binary installer; the skill is instruction/script-based and includes an install.yaml describing intended install paths. No remote fetches or archive extracts are present in the provided files — low install risk. Note: the manifest suggests files will be installed under an npm-global path, but no automated installer is included.
Credentials
The skill requests no credentials or special environment variables by default. It does use/mention configurable env vars (PC_ASSISTANT_OUTPUT_DIR, PC_ASSISTANT_CLEANUP, etc.) which are proportionate. However, the script reads sensitive files (SSH authorized_keys, known_hosts, /var/log/auth.log, user directories and program files under WSL) — this access is explainable for a healthcheck but is sensitive and should be explicitly accepted by the user before running or scheduling.
Persistence & Privilege
always is false and the skill does not request privileged permanent presence. The scheduler will create and optionally delete files in the configured output directory; cleanup is opt-in (PC_ASSISTANT_CLEANUP). The skill does not modify other skills or system-wide agent configuration.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install pc-assistant-fixed
  3. After installation, invoke the skill by name or use /pc-assistant-fixed
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of pc-assistant: cross-platform PC diagnostics and healthcheck skill. - Runs comprehensive, read-only healthchecks on Windows, macOS, and Linux systems - Provides detailed reports including actionable recommendations for detected issues - Supports scheduled and automated healthchecks via cron/scheduler script - Outputs human-readable and JSON reports with system overview, hardware, software, security, and performance stats - Customizable output location and automatic cleanup of old reports via environment variables or config file
Metadata
Slug pc-assistant-fixed
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Pc Assistant Fixed?

PC healthcheck and diagnostics with detailed system information and actionable recommendations. Works on Windows, macOS, and Linux. Read-only system diagnost... It is an AI Agent Skill for Claude Code / OpenClaw, with 240 downloads so far.

How do I install Pc Assistant Fixed?

Run "/install pc-assistant-fixed" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Pc Assistant Fixed free?

Yes, Pc Assistant Fixed is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Pc Assistant Fixed support?

Pc Assistant Fixed is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Pc Assistant Fixed?

It is built and maintained by Ning (@ningtoba); the current version is v1.0.0.

💬 Comments