chencore/tweet-skills

This page was translated automatically. Technical details may differ from the original. View the original

77 stars Quality 43 Trend 18

X Tweet Skills for OpenClaw - A skill pack for automatic tweet creation

Overview

An automated tweet creation skill pack based on OpenClaw on the X platform. Covers the complete pipeline from topic collection to tweet delivery. Each tweet must meet at least 2 of the following: - Data: specific numbers (e.g., 42.5 points, 3 months) - Judgment: unique viewpoint, not common sense - Actionable: tools/methods/steps - Memorable point: summarizing the entire text in one sentence tweet-polish uses a three-level check: - : Automatically replace AI expressions (therefore, → so) - : Detect the proportion of complete sentences; if it exceeds 60%, break sentences - : Create paragraph length discrepancies Copy the folder in the skills/ directory to ~/.openclaw/skills/ to do so: Topic selection → information research → viewpoint generation → post writing → conversational rewriting → title optimization → closing hook → delivery 1. → Automatically collect today's hot topics 2. → Gather materials 3. → 10 viewpoints from different perspectives 4. → 10 styles of posts 5. → Rewrite into a real speaking style 6. → 5 directions of headlines 7. → 5 types of ending conversions - OpenClaw runtime environment - web_fetch / agent-reach (for information collection)

Overview (machine translated)

An automated tweet creation skill pack based on OpenClaw on the X platform. Covers the complete pipeline from topic collection to tweet delivery.

Core Goal: Write 400-600 words of tweets that people won’t want to save.

Skill List

Skills	Functions	Word count/traits	Trigger words
hot-topics	Hot topic collection	10-20 topics	“Collect hot topics” “Find topics”
topic-research	Topic Selection and Research, Collecting Multi-Source Materials	500-800 Words Structured Materials	“Research XX” “Collecting Information”
opinion-generator	Generate unique viewpoints from 10 perspectives	Includes examples and counterexamples	“Generate viewpoints” “Refine perspectives”
tweet-writer	10 Styles of Recommendation Writing	400-600 words, including reasons for collecting	“Write Tweets” “Tweet”
tweet-polish	Colloquial rewriting to eliminate AI flavor	Three-level detection system, AI-flavored scoring	“Colloquial” “polish”
title-optimizer	Title optimization in 5 directions	No more than 20 characters per direction	“Title Setup” “Title Optimization”
tweet-closing	5 Types of Closing Conversion Hooks	2-3 Alternatives/Types	“Write Closing” “Add Hooks”
tweet-pipeline	Main Control Pipeline: One-Click Full Workflow	400-600 Words Finished Product	“Assembly Line” “One-Click Recommendation”

Core Features

Collectible value design

Each tweet must meet at least 2 of the following:

Data available: specific numbers (e.g., 42.5 points, 3 months)
Judgment: unique perspectives, not common sense
Actionable: tools/methods/steps
Memorable points: summarize the entire text in one sentence

Go to AI-scented systems

Tweet-polish uses three-level checking:

L1 Vocabulary: Automatically replaces AI expressions (therefore→so)
L2 Sentence Structure: Detects the proportion of complete sentences; if it exceeds 60%, it is broken down
L3 Rhythm: Create segment length variations

Target: AI Flavor Score ≤ 4

Unify word count standards

Stage	Word Count Requirement
tweet-writer output	400-600 words
tweet-polish Rewritten	400-600 words
topic-research Material	500-800 words
title-optimizer Title	No more than 20 characters
tweet-closing Conclusion	No more than 2 sentences

Installation

Copy the folder from the skills/ directory to ~/.openclaw/skills/:

Usage Process

Method One: Assembly Line Mode (Recommended)

Just say ‘Write XX topic on an assembly line’ and automatically follow the full process:

This page was translated automatically. Technical details may differ from the original.

Recommended Tools

Try a different keyword or remove a filter.

Install

npx skillfish add chencore/tweet-skills

New to MCP? Start with the mcpskill.org guide →