Control Microsoft Edge browser to fetch web pages and extract content. Use when the user wants to access a specific URL using Edge browser instead of Chrome,...

This skill appears coherent and implements exactly what it claims: using Playwright to control Edge/Chromium and return rendered page content as JSON. Before installing, note: (1) SKILL.md requires you to run 'pip install playwright' and 'playwright install chromium' — Playwright will download browser binaries to disk. (2) The script executes pages' JavaScript (headless browser), so visiting untrusted or internal URLs can cause network requests or trigger side effects on the target site; do not use it to fetch internal-only or sensitive endpoints (localhost, internal APIs) unless you understand the risks. (3) The code does not exfiltrate data to external servers beyond returning the page content in stdout/file, but that content could itself contain sensitive data; consider running the skill in a sandboxed environment or with limited network access. (4) The skill source is listed as unknown — if provenance matters, verify the publisher before installing. If you want stronger assurance, ask the publisher for signed releases or run the script in an isolated container and inspect outputs during test runs.

Type: OpenClaw Skill Name: edge-browser Version: 1.0.0 The skill is classified as suspicious due to two main vulnerabilities: 1) The `scripts/fetch.py` allows writing fetched web content to an arbitrary file path specified by the `--output` argument. If an attacker can control this path, it could lead to arbitrary file write, potentially overwriting sensitive files. 2) The `SKILL.md` instructs the OpenClaw agent to execute `scripts/fetch.py` with a user-provided URL. If the agent does not properly sanitize this URL before passing it to the shell command, it could be vulnerable to shell injection. There is no evidence of intentional malicious behavior such as data exfiltration, persistence, or remote code execution within the skill's code or instructions.