Installable skillMIT

LinkedIn Post Generator Skill

Generate high-quality LinkedIn posts for GEO thought leadership. Use when asked to write LinkedIn content, create social media posts, or generate content about SEO/GEO topics.

BY guia-matthieuUPDATED Oct 28, 2026
INTENDED FOR◈OpenClawInstall in a listed agent.

instructions.md

---
name: linkedin-post-1ed6f948718b
description: Generate high-quality LinkedIn posts for GEO thought leadership. Use when asked to write LinkedIn content, create social media posts, or generate content about SEO/GEO topics.
license: MIT
metadata:
  author: ClawFu
  version: 1.0.0
  mcp-server: "@clawfu/mcp-skills"
---

# LinkedIn Post Generator Skill

Generate high-quality LinkedIn posts for GEO thought leadership.

## Trigger
Use this skill when asked to write a LinkedIn post, or with `/linkedin-post`

## Process

### Step 1: Understand the Request
Ask if not provided:
- **Topic**: What's the main subject?
- **Pillar**: education | case_study | insight | behind_scenes
- **Context**: Any specific data, audit results, or angle?

### Step 2: Choose Structure by Pillar

#### Education Post
```
[HOOK - 1 ligne choc]

[CONTEXT - Pourquoi c'est important]

[CORE INSIGHT - 3-5 points clés]

[TAKEAWAY - Ce qu'il faut retenir]

[CTA - Question ou appel]
```

#### Case Study Post
```
[HOOK - Le résultat choc]

[DATA - Les chiffres clés]

[ANALYSIS - Ce que ça révèle]

[LESSON - L'apprentissage]

[CTA - Proposer un audit]
```

#### Insight Post
```
[HOOK - L'actualité/tendance]

[CONTEXT - Ce qui s'est passé]

[ANALYSIS - Ce que ça signifie]

[IMPLICATION - L'impact pour l'audience]

[CTA - Question ou réflexion]
```

#### Behind the Scenes Post
```
[HOOK - Story opener]

[STORY - Ce qui s'est passé]

[LESSON - Ce que j'ai appris]

[ASK - Question à l'audience]
```

### Step 3: Apply the 3 Questions Test

Before finalizing, check:

1. **VISUALIZABLE?** Can the reader "see" it?
   - ❌ "Solution innovante" → ✅ "Votre marque citée par ChatGPT"

2. **FALSIFIABLE?** Is it verifiable?
   - ❌ "Excellent service" → ✅ "Score de 67% sur Perplexity"

3. **UNIQUE?** Could a competitor say this?
   - ❌ "Expert SEO" → ✅ "J'ai audité 50 sites avec mon outil GEOTOOL"

### Step 4: Hook Bank (Use for Inspiration)

**Counterintuitive:**
- Votre SEO est parfait. Mais les IA ne vous voient pas.
- Le meilleur contenu ne suffit plus.
- Google vous adore. ChatGPT vous ignore.

**Surprising Data:**
- 23% de visibilité sur GPT-4. 67% sur Perplexity.
- 1 entreprise sur 10 apparaît dans les réponses ChatGPT.

**Question:**
- ChatGPT recommande-t-il votre marque ?
- Savez-vous ce que les IA disent de vous ?

**Bold Claim:**
- Le SEO tel que vous le connaissez est mort.
- Si les IA ne vous citent pas, vous n'existez plus.

**Story Opener:**
- J'ai audité la visibilité IA d'un client. Le résultat m'a choqué.
- La semaine dernière, j'ai fait 5 audits gratuits.

### Step 5: Format Rules

- **Language**: French, professional but accessible
- **Length**: 150-250 words
- **Paragraphs**: 1-2 lines max
- **Line breaks**: Use liberally for readability
- **Emojis**: 1-2 max if really relevant (not required)
- **Hashtags**: 3-5 at the end (optional)
- **CTA**: Always end with question or call to action

### Step 6: Output

Provide the post directly, ready to copy-paste to LinkedIn.
No introduction, no "Here's your post:", just the content.

---

## Example Output

```
Votre site est invisible pour ChatGPT.

Et ce n'est pas un problème de SEO.

C'est un problème de GEO.

Voici 3 raisons :

1. Les LLMs ne crawlent pas comme Google
→ Ils utilisent des bases de connaissances pré-entraînées

2. Le format compte plus que les mots-clés
→ Structure claire, données factuelles, sources citables

3. L'autorité se mesure différemment
→ Pas de PageRank, mais des citations croisées

Ce que ça change pour vous :

Un bon classement Google ≠ visibilité IA.

Il faut optimiser pour les deux.

Vous avez déjà vérifié si ChatGPT recommande votre marque ?

#GEO #IA #SEO #Marketing
```

---

## Integration with GEOTOOL

If user provides audit data, use real numbers:
- Global score
- Scores by model (GPT-4o, Claude, Perplexity, Gemini)
- Mention rates
- Recommendations

Turn audit results into compelling case study posts.

---

## What Claude Does vs What You Decide

| Claude handles | You provide |
|---------------|-------------|
| Applying post structure by pillar | Topic and angle selection |
| Writing hooks from hook bank | Approval of final hook |
| Formatting for LinkedIn readability | Voice authenticity check |
| Generating data-driven case studies | Real audit/data numbers |
| Running 3 Questions Test | Final posting decision |

---

## Skill Boundaries

### This skill excels for:
- GEO/SEO thought leadership content
- Data-driven case study posts
- Educational explainer posts
- Behind-the-scenes insights

### This skill is NOT ideal for:
- Personal/emotional posts → Need authentic human voice
- Controversial opinions → Requires your conviction
- Company announcements → Different format needed

---

## Skill Metadata

```yaml
name: linkedin-post
category: social
version: 2.0
author: GUIA
source_expert: GEO Thought Leadership Framework
difficulty: beginner
mode: cyborg
tags: [linkedin, social-media, content, geo, thought-leadership]
created: 2026-02-03
updated: 2026-02-03
```
COMPATIBILITY & REFERENCES

Platform guidance

Imported from the listed AgentSkill.sh category. Complete supplied package files and notices are retained. Required external tools and credentials are not installed. No bundled code was executed during import. Individual agent task behavior has not been tested.