Copy all TOML content definitions from librefang core repo: - 33 agent definitions (agents/*/agent.toml) - 14 hand definitions with docs (hands/*/HAND.toml + SKILL.md) - 25 integration templates (integrations/*.toml) - 2 example skill definitions (skills/custom-skill-*) - 1 new provider (providers/vertex-ai.toml) Part of the framework-vs-content registry split (RFC v0.7).
6.9 KiB
6.9 KiB
name, version, author, description, tags, runtime
| name | version | author | description | tags | runtime | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| linkedin-hand-skill | 1.0.0 | LibreFang | Expert knowledge for AI LinkedIn management -- API reference, content strategy, networking playbook, and professional engagement best practices |
|
prompt_only |
LinkedIn Management Expert Knowledge
LinkedIn API Reference
Authentication
LinkedIn API uses OAuth 2.0 with bearer tokens.
Bearer Token:
Authorization: Bearer $LINKEDIN_ACCESS_TOKEN
Core Endpoints
Get authenticated user info:
curl -s -H "Authorization: Bearer $LINKEDIN_ACCESS_TOKEN" \
-H "LinkedIn-Version: 202405" \
"https://api.linkedin.com/rest/userinfo"
Create a text post (Posts API):
curl -s -X POST "https://api.linkedin.com/rest/posts" \
-H "Authorization: Bearer $LINKEDIN_ACCESS_TOKEN" \
-H "Content-Type: application/json" \
-H "LinkedIn-Version: 202405" \
-d '{
"author": "urn:li:person:MEMBER_ID",
"lifecycleState": "PUBLISHED",
"commentary": "Your post content here",
"visibility": "PUBLIC",
"distribution": {
"feedDistribution": "MAIN_FEED"
}
}'
Comment on a post:
curl -s -X POST "https://api.linkedin.com/rest/socialActions/URN/comments" \
-H "Authorization: Bearer $LINKEDIN_ACCESS_TOKEN" \
-H "Content-Type: application/json" \
-H "LinkedIn-Version: 202405" \
-d '{
"actor": "urn:li:person:MEMBER_ID",
"message": {"text": "Your comment here"}
}'
Like a post:
curl -s -X POST "https://api.linkedin.com/rest/socialActions/URN/likes" \
-H "Authorization: Bearer $LINKEDIN_ACCESS_TOKEN" \
-H "Content-Type: application/json" \
-H "LinkedIn-Version: 202405" \
-d '{
"actor": "urn:li:person:MEMBER_ID"
}'
Rate Limits
| Endpoint | Limit | Window |
|---|---|---|
| Posts | 25 posts | 24 hours |
| Comments | 10 comments | 1 minute |
| Likes | 20 likes | 1 minute |
| API calls (general) | 100 requests | 1 day |
LinkedIn Content Strategy
The LinkedIn Algorithm (2024-2025)
Key factors that affect reach:
- Dwell time: How long people spend reading your post
- Early engagement: Comments in the first hour boost distribution
- Meaningful comments: Long comments signal quality content
- No external links: Posts with links get 40-50% less reach
- Personal stories: Narrative content outperforms promotional content
Content Pillars
Define 3-4 content pillars:
Example for a tech leader:
Pillar 1: Engineering Leadership (40%)
Pillar 2: Industry Trends & Analysis (30%)
Pillar 3: Career Growth & Mentoring (20%)
Pillar 4: Personal Lessons (10%)
Post Formats That Work
| Format | Avg Engagement | Best For |
|---|---|---|
| Personal story with lesson | High | Connection, authenticity |
| Contrarian take | High | Discussion, visibility |
| Step-by-step guide | Medium-High | Authority, saves |
| Data + insight | Medium | Credibility |
| Question/poll | Medium | Engagement |
| Industry news + analysis | Medium | Thought leadership |
Optimal Posting Times (UTC-based)
| Day | Best Times | Why |
|---|---|---|
| Tuesday | 8-10 AM | Peak professional engagement |
| Wednesday | 8-10 AM | Mid-week content consumption |
| Thursday | 8-10 AM, 12 PM | Second-best engagement day |
| Monday | 8-10 AM | Start of work week |
| Friday | 8-9 AM only | Engagement drops after morning |
| Weekend | Avoid | 60-70% lower engagement |
Post Writing Best Practices
The Hook (First 2 Lines)
The first 2 lines appear before the "see more" fold. They must compel a click.
Hooks that work:
- Bold statement: "I fired my best employee last week. Here's why it was the right call."
- Surprising data: "Only 3% of engineering managers do this. It changes everything."
- Confession: "I made a $500K mistake in my first year as CTO."
- Question: "Why do 90% of digital transformations fail?"
- Contrarian: "Unpopular opinion: Stand-ups are a waste of time."
Writing Rules
- One idea per post -- don't try to cover everything
- Short paragraphs -- 1-2 sentences max, lots of white space
- Use line breaks -- make it scannable
- End with a question -- drives comments which boost reach
- No links in the post body -- put links in the first comment
- 3-5 relevant hashtags -- at the bottom of the post
- 1000-1300 characters -- sweet spot for engagement
- Be authentic -- personal stories outperform corporate speak
Comment Strategy
When commenting on others' posts:
- Add a new perspective or data point
- Share a relevant personal experience
- Ask a thoughtful follow-up question
- Keep comments 2-4 sentences (meaningful but concise)
- Avoid generic comments ("Great post!", "Thanks for sharing!")
Networking Best Practices
Connection Requests
- Always add a personal note (not the default message)
- Reference something specific (their content, mutual connection, shared interest)
- Keep it under 300 characters
- Don't pitch in the connection request
Relationship Building
- Consistently engage with connections' content before asking for anything
- Share others' content with genuine commentary
- Celebrate connections' achievements publicly
- Offer help or resources without expecting anything in return
Safety & Compliance
Content Guidelines
NEVER post:
- Confidential business information
- Discriminatory or offensive content
- False credentials or experience claims
- Defamatory statements about competitors or individuals
- Content that violates LinkedIn's Professional Community Policies
- Misleading data or fabricated statistics
Content Moderation Rules
Before posting any content, classify it:
Auto-REJECT (never post):
- Content containing hate speech, discrimination, or harassment
- Unverified claims about competitors or individuals
- Confidential or proprietary business information
- Content that could be interpreted as financial or legal advice
- Anything with profanity or inappropriate language
- Political or religious debate content
Flag for REVIEW (queue for human approval):
- Controversial industry opinions or contrarian takes
- Content mentioning specific companies or individuals by name
- Posts discussing salary, compensation, or workplace issues
- Content referencing current news events
- Posts with strong emotional tone or personal vulnerability
Safe to POST (can auto-publish if approval_mode is off):
- Educational how-to content and professional tips
- Industry trend analysis with cited sources
- Career development advice and frameworks
- Engagement posts (professional questions, polls)
- Celebration of team or industry achievements
Professional Standards
- Maintain professional tone even in casual posts
- Fact-check all claims and statistics
- Credit sources and tag collaborators
- Disclose affiliations when discussing products or services
- Respect intellectual property and copyright