gooseworks-ai/create-linkedin-content
> Draft voice-tuned LinkedIn post variants from a free-form brief. Reads a personal voice guide (generated via generate-voice-guide), produces 2–5 variants with distinct framings, applies LinkedIn-specific defaults (arrow bullets, "why this matters" beat, 150–500 words), and self-checks against the voice guide's banned phrases before returning.
npx skills add https://github.com/gooseworks-ai/goose-skills --skill create-linkedin-content
Draft LinkedIn post variants that sound like a real human practitioner, not a LinkedIn thought-leader persona. Reads a voice guide the user has already generated (or prompts to create one), produces multiple framings of the same idea, and saves each variant as its own markdown file with frontmatter.
This is an agent-executed skill — the agent does the drafting and self-check inline. No Python script.
Mirrors create-x-content structurally; differences are marked ★.
/create-linkedin-content --brief "New open-source CLI that turns Figma files into React components. Called figma2react. Free, MIT licensed."
Or interactively:
/create-linkedin-content
| Flag | Required | Default |
|------|----------|---------|
| --brief | Yes (asked interactively if missing) | — |
| --variants | No | Skill decides based on brief richness (2–5) |
| --voice-guide | No | Resolved via chain below |
| --output | No | ./content/YYYY-MM-DD-<topic-slug>/ |
| --topic | No | Derived from brief |
Resolve in this order, stop at first hit:
--voice-guide <path> flag~/.goose-skills/config.json → voice_guides.linkedin~/.goose-skills/voice-guides/voice-linkedin.md (default path)/generate-voice-guide --platforms linkedin now to create one (recommended)Apply these unless the voice guide explicitly says otherwise:
→) for workflow steps — this is LinkedIn-native1️⃣ 2️⃣ 3️⃣ allowed sparingly for installation/quickstart type postsSame as create-x-content Phase 1 — load voice guide, extract banned phrases, hook patterns, format rules, dos/don'ts, examples.
Same 2–5 decision rubric. LinkedIn briefs often warrant fewer variants than X because LinkedIn posts are longer and the audience expects more narrative coherence — 3 strong framings usually beats 5.
Framings tuned for LinkedIn (subset of the X framings, plus a few LinkedIn-native ones):
| Framing | Structure | When to use |
|---------|-----------|-------------|
| builder-story ★ | "I built X. Here's why. Here's what I learned." | LinkedIn-native — builder stories outperform tactical posts here |
| product-launch ★ | Clear hook + 2-step quickstart + link | New tool/product announcements |
| problem-first | Open with the pain, then solution | Universal |
| ecosystem-map | Curated list of N related tools/companies | Landscape posts |
| contrarian-insight | "Your X might be worth more than Y" | Opinion pieces with a fresh take |
| workflow-tutorial | Arrow-bulleted step-by-step, with numbers | Tactical how-tos (denser than X version) |
Same 4 checks as create-x-content:
Plus LinkedIn-specific:
→ style (or numbered emojis), matching the LinkedIn convention?File naming (the linkedin- prefix distinguishes from X variants when they live in the same folder):
linkedin-<letter>-<framing-slug>.md
Examples: linkedin-a-builder-story.md, linkedin-b-product-launch.md.
Frontmatter:
---
id: <topic-slug>-li-<letter>
platform: linkedin
format: standard | long
topic: <slug>
framing: <framing-slug>
status: draft
---
Print output directory, file list, framing summary, suggested next step.
<output>/linkedin-<letter>-<framing>.md per variantBuilder story brief:
/create-linkedin-content --brief "Built a CRM for my AI agents using just markdown files and Claude Code. No Salesforce, no HubSpot. Works better than both."
→ 3 variants (builder-story, contrarian-insight, workflow-tutorial)
Launch brief:
/create-linkedin-content --brief "Launching goose-aeo — open-source CLI that measures your brand's visibility on AI search engines like ChatGPT, Perplexity, Gemini. npm install, three commands to run."
→ 3 variants (product-launch, problem-first, workflow-tutorial)
generate-voice-guide skill (for creating one when missing)Take gooseworks-ai/create-linkedin-content from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.
The instructions reference npm.
Without those the skill loads but fails at the first command.