x-agent

Plan, monitor, draft, and safely automate X (Twitter) account workflows with configurable guardrails and phased operational modes.

This skill looks coherent and focused on X/Twitter workflows, with sensible safety guidance (monitor-only start, explicit approvals, kill-switch). Things to check before installing or using it: 1) Confirm where and how you will supply X API credentials—the SKILL.md says to configure them locally but the skill metadata doesn't declare required env vars, so plan to store them in your environment (not in chat). 2) Start in monitor-only mode and test the approval flow before enabling any automation. 3) Verify playbooks and hard caps (max posts/day, quiet hours, kill switch) match your risk appetite. 4) Because the skill source/homepage is unknown, prefer using it only with accounts where you control the credentials and avoid granting broad posting rights until you trust the workflow. If you want a tighter assessment, provide any additional metadata or an updated manifest that declares the exact env vars the skill expects (e.g., X_API_KEY, X_API_SECRET, etc.).

Type: OpenClaw Skill Name: x-agent Version: 0.1.0 The OpenClaw AgentSkills bundle for the 'x-agent' skill is classified as benign. The `SKILL.md` documentation clearly outlines a phased approach for managing X (Twitter) operations, emphasizing strict guardrails, explicit user approval for posting, and disabling automation by default. It also instructs users to configure API credentials locally and 'never in chat,' which is a good security practice. There is no evidence of prompt injection attempts, data exfiltration, malicious command execution, persistence mechanisms, or other indicators of intentional harmful behavior. The described capabilities are aligned with the stated purpose and include significant safety measures.