← Back to Skills Marketplace
Skill Review
by
Oliver Drobnik
· GitHub ↗
· v0.2.4
1208
Downloads
2
Stars
4
Active Installs
8
Versions
Install in OpenClaw
/install skill-review
Description
Scrape ClawHub skill pages for Security Scan (VirusTotal/OpenClaw) + Runtime Requirements + Comments for all of Oliver's local skills, and write a markdown r...
Usage Guidance
This skill appears to do what it says: it scans your local skills' SKILL.md to form ClawHub slugs, loads each ClawHub page in a headless browser (Playwright), extracts security scan summaries, and uses your VirusTotal API key (or ~/.openclaw/.env fallback) to enrich results. Before running: inspect scripts/skill_review.py and suppressions.json yourself (they're bundled), decide whether you are comfortable providing your VIRUSTOTAL_API_KEY (the script will use it to query VT and may cause VT requests tied to your key), and be aware that installing Playwright will download Chromium/brand binaries from upstream. If you don't want network access for a run, don't execute the script — it visits external sites (clawhub.ai and VirusTotal) by design.
Capability Analysis
Type: OpenClaw Skill
Name: skill-review
Version: 0.2.4
The skill is designed to scrape security scan results from ClawHub and VirusTotal for local skills. It accesses local skill definitions under `~/Developer/Skills` (user-configurable), makes network requests to `clawhub.ai` and `virustotal.com`, and writes a report to `/tmp/`. The `VIRUSTOTAL_API_KEY` is handled securely via environment variables or `~/.openclaw/.env`. All observed behaviors, including file system access and network calls, are directly aligned with its stated purpose. There is no evidence of data exfiltration, malicious execution, persistence, or prompt injection attempts against the agent in `SKILL.md`.
Capability Assessment
Purpose & Capability
The skill says it scrapes ClawHub pages and gathers Security Scan (VirusTotal/OpenClaw) info and runtime requirements for local skills. It requires python3 and VIRUSTOTAL_API_KEY and uses Playwright to render client-side content — all of which are coherent with that goal.
Instruction Scope
Runtime instructions and included script enumerate local skill folders (SKILL.md), derive slugs, visit https://clawhub.ai/<owner>/<slug>, expand page details, and extract Security Scan / Runtime / Comments. The script only reads SKILL.md, optional slug-map, and a local suppressions.json (next to the script or in parent) and may read ~/.openclaw/.env as a fallback for the VT key. It does not attempt to read unrelated system files or credentials.
Install Mechanism
No install spec is provided (instruction-only with bundled script). The script depends on Playwright (Python) and therefore the user will likely install the playwright package and download Chromium via 'playwright install chromium' — this downloads a browser binary from upstream, which is expected but worth noting as an external binary download.
Credentials
Only VIRUSTOTAL_API_KEY is required (declared). The script legitimately uses the VT key to query VirusTotal and implements client-side rate limiting. The only other env-read behavior is a fallback to ~/.openclaw/.env to find the same key; no unrelated secrets or credentials are requested.
Persistence & Privilege
The skill is user-invocable and not set to always: true. It writes an output markdown to the specified --out path (example /tmp/) and may create Playwright browser artifacts if you install Playwright, but it does not persistently modify other skills or system-wide agent settings.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install skill-review - After installation, invoke the skill by name or use
/skill-review - Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.2.4
fix: use /Users/oliver/clawd for workspace root to preserve symlink paths
v0.2.3
Fix env var declarations: move into requires.env as array (ClawHub metadata recognition)
v0.2.2
Declare missing dependencies, env vars, and sensitivity in metadata
v0.2.1
Rename .clawdhubignore to .clawhubignore
v0.2.0
Refactor: move setup/prerequisites to SETUP.md, keep SKILL.md lean
v0.1.2
MD formatting: avoid duplicating Security Scan raw fallback when DOM scrape succeeds.
v0.1.1
Scraper improvements: wait for client-side render; extract Security Scan from DOM (VT link + OpenClaw details); add --only.
v0.1.0
Initial release: scrape ClawHub Security Scan + Runtime requirements + Comments for all local skills (Playwright).
Metadata
Frequently Asked Questions
What is Skill Review?
Scrape ClawHub skill pages for Security Scan (VirusTotal/OpenClaw) + Runtime Requirements + Comments for all of Oliver's local skills, and write a markdown r... It is an AI Agent Skill for Claude Code / OpenClaw, with 1208 downloads so far.
How do I install Skill Review?
Run "/install skill-review" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Skill Review free?
Yes, Skill Review is completely free (open-source). You can download, install and use it at no cost.
Which platforms does Skill Review support?
Skill Review is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Skill Review?
It is built and maintained by Oliver Drobnik (@odrobnik); the current version is v0.2.4.
More Skills