Fetch live exchange rates between any currency pairs from XE.com. Use when: user asks about currency conversion, exchange rates, forex rates, or converting a...

This skill appears to perform XE scraping with a Playwright browser and an API fallback, but there are red flags you should address before installing: - The script hard-codes a CDP websocket URL including a token (ws://localhost:7002?token=...). Ask the author why a local CDP endpoint and token are required, where that token comes from, and why it isn't declared as a required service or env var. - A local CDP connection can give the skill access to any browser context on that service (cookies, sessions, localStorage). Only run this skill in a controlled/sandboxed environment where no sensitive browser sessions are exposed. - The manifest does not declare dependencies (playwright-core, Node fetch availability) or installation steps. Confirm platform provides Playwright and a compatible browser/CDP, or update the skill to declare/install them. - Prefer removing embedded tokens; if a token is required, it should be supplied via a documented env var or config (and treated as a secret). If you need this functionality, ask the publisher for an updated manifest that documents required services/dependencies and explains the CDP endpoint/token. If you cannot get that, run the skill in an isolated environment or decline installation.

Type: OpenClaw Skill Name: exchange-rates Version: 1.0.1 The skill is classified as suspicious due to a hardcoded CDP token (`ec546a08aed110e96f64cc645bdb58fa8829a63349d6ae53`) in `scripts/xe-rate.mjs` which is a security weakness. Additionally, the `scripts/xe-rate.mjs` script directly uses `process.argv` inputs (currency codes, amount) without explicit sanitization, which, depending on the agent's command execution mechanism, could lead to shell injection vulnerabilities. These are significant security flaws, but there is no clear evidence of intentional malicious behavior like data exfiltration or backdoor installation.