review-manager

고객사 리뷰 수집·자동답글·알림·리포트 통합 관리. 네이버플레이스/구글/배민/쿠팡 리뷰 모니터링 + 감성분석 + 경쟁사 비교

This skill appears to do what it says, but review and cautious deployment are advised: 1) It will create and update files under ~/.openclaw/workspace/skills/review-manager/data — confirm you are comfortable with that location. 2) Automated Discord notifications are sent via the host's openclaw message CLI (not by shipping Discord tokens); check that the messaging tool has appropriate permissions and that the target channel ID in config.json is correct. 3) The project currently uses mock collectors and notes that real scraping may require browser automation and cookie reuse — avoid storing login cookies or credentials in plaintext; if you implement login automation, treat those secrets carefully. 4) The --apply auto-reply functionality is not implemented yet (scripts generate replies and save previews only). 5) If you plan to run these scripts on a schedule, run them in a controlled environment and ensure scraping frequency complies with each platform's terms to avoid IP blocking or legal issues. If you want higher assurance, run the code in a sandbox/container and inspect the openclaw message tool behavior before enabling automated notifications.

Type: OpenClaw Skill Name: review-manager Version: 1.0.0 The skill is classified as suspicious due to a shell injection vulnerability present in `scripts/check-negative.js` and `scripts/weekly-report.js`. Both scripts construct a shell command using `child_process.execSync` to send Discord alerts. The `discordChannelId` from `config.json` is directly interpolated into this command string without proper shell escaping, allowing for arbitrary command execution if a malicious value is configured for `discordChannelId` (e.g., `1234567890; rm -rf / #`). While the `message` content is sanitized, the `channelId` is not, posing a significant risk.