LinkedIn Post Writer skill and the Yalc Framework
The single highest-leverage content skill if you post on LinkedIn weekly. Voice-calibrated, anti-detection rules baked in, written from operator memory of what works.
Say this to fire the LinkedIn Post Writer skill
Any of these natural language phrases activates the skill inside Claude Code.
LinkedIn Post Writer, plainly
The LinkedIn Post Writer skill drafts posts in Othmane Khadri's voice for the Yalc and Earleads brands. Voice calibration comes from a corpus of past winning posts plus 10 explicit rules captured from live edits ("don't start with I", "no em dashes", "no Here's the X", etc.).
The skill applies two voice profiles depending on topic: Voice 1 (GTM Intelligence) for posts a CMO would care about, Voice 2 (AI-Native Builder) for posts an engineer would care about. The split rule is "Would a CMO care more, or would an engineer care more?"
Position in the GTM operating system
The LinkedIn Post Writer sits at the **draft** node for LinkedIn content. It produces a post draft in the right voice profile, with the rules audit applied (no em dashes, no AI tells, no throat-clearing intros).
Output is a draft ready for human review. The post never goes live without Othmane's eyes on it. The skill's value is in producing a 90 percent draft that needs minimal editing, not a 50 percent draft that needs full rewrite.
Running the LinkedIn Post Writer skill end to end
Workflow position
The LinkedIn content drafter. Yalc invokes this skill when the user wants a post written. Output is a draft; publishing is manual through LinkedIn's UI or via the Unipile API.
Required inputs
- → Topic or thesis (the core insight or story to share)
- → Voice profile (auto-detected via the CMO vs engineer rule, or explicit override)
- → [object Object]
- → [object Object]
Outputs
- → Post draft in markdown ready for LinkedIn paste
- → Hook line (first 2 lines, since LinkedIn shows only those before "see more")
- → Suggested hashtags (or none, depending on post archetype)
- → [object Object]
Chaining recommendations
Anti patterns to avoid
Pros, cons, who it's for
Pros
- Two voice profiles cover the full range of post archetypes
- 10 explicit rules from live edits prevent common AI tells
- Hook-aware (optimizes for LinkedIn's "see more" cutoff)
- Optional visual proposition output for posts that benefit from imagery
- Othmane's calibrated voice (not generic LinkedIn marketing slop)
Cons
- Specific to Othmane's voice. For other operators, fork the rules and corpus.
- Doesn't auto-publish. Manual paste or separate tool required.
- Single-language (English). For non-English posts, voice calibration is weaker.
- Long-form posts (above 1500 chars) sometimes lose voice consistency mid-post
Who it's for
- Othmane Khadri and the Yalc/Earleads brand voices
- Operators who post on LinkedIn weekly and want voice consistency
- Founders building thought leadership through LinkedIn content
What this skill expects to find
Voice corpus lives at `02_Areas/Marketing/Brand_Voice/` (Voice 1 and Voice 2 profile files). Live-edit rules at `memory/linkedin-writing-rules.md`. The skill reads both at runtime. Updating either changes the output style for all future posts.
The LinkedIn Post Writer ecosystem inside Yalc
Skills that overlap
Frequently asked
How does the skill choose between Voice 1 and Voice 2?
Auto-detection based on topic. If the angle leans GTM strategy, market positioning, or commercial framing, Voice 1 (CMO-care). If it leans engineering, AI-native architecture, or technical experimentation, Voice 2. Override explicitly if needed.
Why no em dashes?
Em dashes are an AI signature pattern. LinkedIn audiences (especially the AI-savvy ones Othmane targets) recognize them as LLM-generated. The skill grep-checks output before returning.
Can the skill write threads (multi-part posts)?
Yes. Specify "thread" in the input. The skill produces a sequence of related posts with hook continuity across them.
What's the optimal post length?
Depends on archetype. Insights and hooks land at 600 to 1200 chars. Stories land at 1200 to 1800. Lists land at 800 to 1500. The skill respects the target length you specify.
How is the voice corpus updated?
Manually. When new posts perform well, add them to `02_Areas/Marketing/Brand_Voice/Voice_X_examples.md`. The skill reads the latest file at runtime, so updates take effect immediately.
Does the skill handle hashtags?
Conditionally. Some post archetypes benefit from 2 to 3 targeted hashtags; others (story posts) read better without. The skill recommends based on archetype; you decide.
Get the LinkedIn Post Writer skill
Clone the Yalc skill set, drop in your env, run from your next Claude Code session.
gh repo clone Othmane-Khadri/YALC-the-GTM-operating-system && cp -r YALC-the-GTM-operating-system/.claude/skills/linkedin-post ./.claude/skills/