Your AI executive team on Discord. 7 specialists (engineering, finance, marketing, devops, legal, management, chief of staff) each with its own model and per...

This skill is coherent with its goal (running multiple AI-powered Discord bots) but there are important warnings you should heed before installing: - Metadata mismatch: The package registry entry claims no required env vars or config paths, but SKILL.md and the config template clearly require LLM API credentials and many Discord bot tokens and reference ~/.clawdbot/clawdbot.json. Do not rely solely on the registry metadata — assume you will need to provide sensitive keys. - Inspect the npm package: 'npm install -g clawdbot' will install code that will run with your user privileges and hold your tokens. Review the package source on the linked GitHub, and preferably inspect the installed code before running it as a service. - Protect credentials: The template stores multiple tokens in ~/.clawdbot/clawdbot.json. Use strict file permissions, consider using a secrets manager or letting the gateway handle auth in memory rather than writing tokens to disk. - Run in a sandbox/least privilege: The skill mentions a sandbox but it is off by default. If possible, enable the containerized, network-isolated sandbox to reduce risk of data exfiltration by agents. If you must run without sandboxing, isolate the server and use dedicated bot accounts with minimal privileges. - Verify Discord settings: The config requires 'groupPolicy': 'open' and enabling privileged intents — that grants bots broader access to messages and members. Use dedicated bot accounts and only enable intents required for functionality. - Service review: Before starting 'systemctl --user start clawdbot-gateway', inspect the service unit (what user, what environment it will load) to ensure it won't inadvertently expose secrets. If you want, I can: (1) list the exact placeholders and sensitive fields from the template for you to audit, (2) suggest a safer workflow for keeping tokens out of disk (e.g., environment injection, secrets manager), or (3) help review the GitHub repo source if you provide its package contents or a link to the npm package contents.

Type: OpenClaw Skill Name: become-ceo Version: 1.5.0 The skill bundle is classified as suspicious due to its reliance on high-privilege system operations and handling of sensitive credentials, which, while plausibly needed for its stated purpose, introduce significant security risks. Specifically, the `SKILL.md` instructs the OpenClaw agent to execute `npm install -g clawdbot` and `systemctl --user start clawdbot-gateway`. These commands involve installing global packages and managing system services, which could lead to Remote Code Execution (RCE) or persistence vulnerabilities if the `clawdbot` package itself were compromised or if the agent were manipulated to install a malicious version. The skill also requires handling sensitive `LLM_API_KEY` and `DISCORD_BOT_TOKEN` credentials, increasing the potential impact of a compromise, despite instructions in `references/AGENTS.md` to 'Never leak private data'.