Usage Monitor

Monitor any service usage dashboard and alert when threshold is reached. User-configurable URL and threshold.

This skill is not malicious, but it is misleading: it advertises automated monitoring and alerting but the provided check.js only (a) loads/validates a local config.json, (b) formats alert messages given usage data, and (c) updates a local usage-log.md. It does not fetch the panelUrl, parse usage percentages, or send messages to any channel. Before installing or relying on this skill: - Inspect config.json and avoid putting secrets or shareable tokens in panelUrl; treat the URL as private. - Understand you'll need an external mechanism (browser automation, HTTP fetch + parser, or an agent capability that implements browser.open/snapshot and message.send) to collect usageData and to deliver alerts. The SKILL.md shows these steps only as pseudocode. - If you plan to enable autonomous agent invocation for end-to-end monitoring, implement or review the code that does network access and message sending, and re-audit it for network endpoints and credential usage. - Run the script in a controlled environment first to confirm behavior and to ensure it doesn't log sensitive URLs to public repos. If you want something that truly auto-scrapes dashboards and posts alerts, request that the author add explicit, auditable code (and list any required credentials) rather than relying on manual/pseudocode steps.

Type: OpenClaw Skill Name: usage-monitor Version: 2.0.0 The 'usage-monitor' skill is a legitimate utility designed to help users monitor service usage dashboards via the OpenClaw agent. The core logic in `check.js` handles configuration validation and local logging to `usage-log.md`, while providing clear instructions for the AI agent to use its built-in browser tools to fetch data from a user-defined URL. No evidence of data exfiltration, malicious execution, or prompt injection attacks was found; the skill operates entirely within its stated purpose.