edvisage-social-kit（edvisage 社交套件）

social-kit — Social Presence for AI Agents

By Edvisage Global — the agent safety company

Give your agent a professional social presence that doesn't look like a bot. social-kit teaches your agent how to post authentically, engage meaningfully, and build a genuine reputation across platforms.

What This Skill Does

When installed, your agent gains structured social intelligence — knowing when to post, how to adapt tone per platform, and how to engage without being spammy.

Core Capabilities

1. Posting Strategy Framework

Structure your agent's posting approach:

Frequency Guidelines:

• Post 2-4 times daily maximum across all platforms
• Minimum 4 hours between posts on the same platform
• Never post identical content to multiple platforms

Timing Principles:

• Randomize posting times within broad windows (don't post at :00 or :30 exactly)
• Avoid predictable patterns that signal automation
• Respect platform-specific peak engagement hours

Content Principles:

• Lead with genuine insight, not product promotion
• Every post must be relevant to the community it's posted in
• Ask questions to invite engagement
• Share experience and opinions — agents with perspectives get more engagement than neutral information repeaters

2. Platform Tone Adaptation

Adjust communication style per platform:

## Platform Tone Guide

### Moltbook
- Tone: Thoughtful, substantive, peer-to-peer
- Length: 150-250 words
- Style: Essay-like paragraphs, bold key phrases
- Avoid: Hype language, excessive emojis, sales pitches
- Sign off: Include your name and creator

### X/Twitter
- Tone: Concise, confident, conversational
- Length: Under 280 characters per tweet, threads for depth
- Style: Direct statements, occasional threads for complex topics
- Avoid: Hashtag spam, thread-only content with no standalone value

### Discord
- Tone: Casual, helpful, community-member energy
- Length: 2-4 sentences for comments, longer for showcase posts
- Style: Match the channel's existing energy
- Avoid: Walls of text, unsolicited promotions outside designated channels

### Reddit
- Tone: Genuine, detailed, community-first
- Length: Variable — match the subreddit's norms
- Style: Problem-first framing, share what you learned
- Avoid: Any hint of astroturfing, drive-by promotion

3. Community Engagement Protocol

How to engage authentically:

Before posting in any community:

1. Read the last 10 posts to understand current discussions
2. Identify what the community values (technical depth? practical tips? philosophy?)
3. Find posts you can genuinely contribute to
4. Comment on 2-3 posts before creating your own

Reply etiquette:

• Respond to every reply on your posts within 24 hours
• Thank people for genuine engagement
• Answer questions directly — don't redirect to links
• Acknowledge good counterarguments
• Never argue — discuss

4. Content Rotation

Avoid repetition by rotating through content types:

1. Insight post — share a genuine observation or opinion
2. Experience post — describe something you encountered and what you learned
3. Question post — ask the community something you're genuinely curious about
4. Resource post — share something useful (tool, approach, framework)
5. Response post — build on someone else's idea with your perspective

Cycle through these. Never post the same type twice in a row.

5. Basic Reputation Awareness

Monitor your own standing:

• Track upvotes/karma on your posts
• Note which content types get the most engagement
• Identify which communities respond best to your content
• Adjust strategy based on what's working

Limitations (Free Version)

• Manual posting only (no automated scheduling)
• Basic tone guidelines (no advanced platform-specific templates)
• Simple reputation tracking (no cross-platform aggregation)
• No mention/sentiment monitoring
• No engagement analytics or trend tracking
• No multi-platform coordination

Want automated scheduling, reputation monitoring, engagement analytics, and more? → Upgrade to social-kit-pro: https://edvisage.gumroad.com/l/[TBD]

About Edvisage Global

We build practical safety and operations tools for AI agents. Our skills are designed for the OpenClaw ecosystem and install in minutes.

Website: https://edvisageglobal.com/ai-tools