Generate LinkedIn Content & Calendars

LinkedIn Content Generator is a full content-creation suite for Claude Code that turns a topic and niche into publish-ready LinkedIn posts, multi-slide carousels, long-form newsletter editions, and 30-day content calendars, wired through a personal reinforcement-learning memory system so output improves as feedback is given. Seven coordinated commands cover the workflow: /generate-post, /generate-carousel, /generate-newsletter, /generate-calendar, /show-memory, /feedback, and /clear-memory. Bundled Python scripts build engineered prompts, inject saved preferences from a local memory.md file, and enforce LinkedIn SEO rules before Claude generates output.

/generate-post produces a single post with a required scroll-stopping 2-line hook, context, a bulleted core-value section (max 7 items), a takeaway, a call to action, and 3-5 hashtags, with tone (professional/storytelling/controversial/educational/motivational) and style (list-based/text-only/storytelling/data-driven/contrarian) parameters. /generate-carousel produces 3-12 numbered slides plus caption in one of five styles (how-to, listicle, myth-busting, framework, story-arc). /generate-newsletter produces a full edition (700/1,200/2,000 word length tiers) with an SEO headline, H2-structured body, takeaways, an action step, and an engagement question. /generate-calendar produces a Markdown table calendar sized to a goal (awareness/engagement/leads/authority/growth), each with its own content-mix strategy.

Enforced LinkedIn SEO rules include a scroll-stopping first line with forbidden generic openers ("In today's...", "Thrilled to announce..."), a pattern-interrupt second line, max 2 sentences per paragraph, Grade-8 reading level, and a hard 5-hashtag limit (1 broad, 2 niche, 1-2 community). The memory loop persists niche, tone, successful hooks, and feedback across sessions via /feedback, so later generations incorporate what specifically worked (e.g. "contrarian hook drove 400% more impressions").

SKILL_SCRIPTS="${HOME}/.claude/skills/linkedin-content-generator/scripts"
python3 "${SKILL_SCRIPTS}/generate_post.py" --topic "..." --niche "..." --tone professional --style list-based

All scripts run locally with no network requests and no credentials read, written, or logged; only memory_manager.py writes, and only to the bundled memory.md.

Use this skill when you need a ready-to-paste LinkedIn post, carousel deck, newsletter edition, or a full month's content calendar with platform-native SEO formatting, or when you want output that adapts to your voice over time via saved feedback.

This skill does not publish to LinkedIn directly - all output is copy-paste ready but requires manual posting. The memory system is file-based and local, not shared across machines without manual sync. It does not auto-schedule or integrate with scheduling tools (Buffer, Hootsuite). The slides parameter is silently clamped to 3-12. Very large memory.md files (500+ entries) may exceed prompt context and need periodic archiving via /clear-memory. Requires Python 3.8+ on PATH; does not work in sandboxed environments without Bash/python3 access.

Inputs: a topic, niche, tone/style parameters per command, and accumulated feedback in memory.md.

Outputs: a publish-ready LinkedIn post, numbered carousel slides with caption, a structured newsletter edition, or a 30-day Markdown content calendar with SEO keywords and format breakdown.

Complements @content-creator (broader cross-platform content), @content-strategy (topic-cluster planning before running the calendar command), @content-marketer (multi-channel campaign planning), @linkedin-automation (programmatic publishing via Composio/Rube MCP after content is generated here), and @linkedin-profile-optimizer (align generated voice with profile branding).

LinkedIn creators, marketers, and founders across any niche who want platform-native, SEO-formatted content generated from a topic and refined over time through saved feedback.