Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
A framework for running an AI agent on X/Twitter with authentic, context-driven engagement rather than scheduled posting. Covers identity setup, organic reply logic, 24-hour deduplication, safety rules for crypto/scam mentions, and a presence loop that posts only when something is worth saying.
Unlike scheduled posting tools, this approach treats silence as a valid output, which keeps an agent's timeline from filling with forced content that reads as obviously automated.
When to use it
- Setting up an AI agent's X voice and bio from existing identity files
- Running a background loop to check X and decide whether to post or stay quiet
- Reviewing mentions to pick which deserve a reply and which to ignore
- Checking recent tweets before posting to enforce the no-repeat-topic rule
- Quietly handling crypto token or CA mentions without amplifying them publicly
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Check X and decide if there's anything worth engaging with right now
- 1Search X for trending topics in the agent's niche using x-research
- 2Fetch the agent's last 10 tweets and scan for topics already covered in the past 24 hours
- 3Browse the mention timeline and filter out bots, spam, and token mentions
- 4Identify one thread with a genuine gap the agent can add something to
- 5Post a concise reply if something qualifies, or return no output if nothing meets the bar
Agent replies to one relevant thread, or returns empty-handed with no post made
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.