evanbaker315/attentionos/linkedin-longform
Claude Code and agent skill by evanbaker315/attentionos
- community
Checking that the SKILL.md file is still in the repository.
Write a LinkedIn long-form post through research, writing, an optional visual and a scored quality rubric (8 overall, no criterion under 7) that the save tool enforces in code. Use for the LinkedIn part of the weekly text-post batch or when the owner asks for a LinkedIn post.
Install
npx vskill@latest install evanbaker315/attentionos/linkedin-longformSKILL.md
LinkedIn long-form
Four steps, in order. Each step's output is the next step's only input.
1. Research
- Start from what the business actually knows:
brand/proof-inventory.yaml,brand/offers.md, recent scripts (scripting_list_recent_scripts) and cited sources (scripting_get_claim_sources). - When the topic needs outside facts, search for primary sources from the last 30 days unless the topic is evergreen. Prefer official announcements, original research and reputable outlets.
- Label every point fact (stated by a credible source, with a detail), inference (your conclusion from facts) or opinion.
- Find one real counterpoint: the strongest reason the thesis might be wrong or incomplete.
- No verified stat? Write "no verified stats" and don't use one. Skip stats already spent in recent scripts.
2. Write
- The first one or two lines carry the hook and land before the "see more" fold (about 210 characters).
- One idea per post. Short paragraphs. One concrete, save-worthy takeaway.
- Every claim traces to the research without being strengthened or generalized; outside single-source claims say whose claim they are (
claim-ledgerskill). - Owner's voice (
brand/voice.md), run through thespoken-humanizerchecks that apply to writing. - 3 to 5 niche hashtags in the hashtags list, never in the text. No links in the body; the link goes in the first comment.
- A CTA that invites a real reply, not engagement bait.
3. Visual (optional)
If the post gets an image, it must show the post's central claim, not just its topic; follow brand/house-style.md; read clearly on a phone; contain no logos or invented brand marks.
4. Score the rubric honestly
Score 1 to 10 each: hook, clarity, usefulness, accuracy, formatting, cta; with a visual also concept_match, house_style, mobile_legibility, technical_quality, compliance.
The bar is 8 overall and no criterion under 7. Accuracy is where real defects hide: check it claim by claim. Below the bar, name the failing criterion, quote the weak line, fix it and score again.
Save with scripting_save_text_posts including the rubric scores. The tool rejects posts under the bar.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingplugin/skills/linkedin-longform/SKILL.md
Popularity
- Repository stars
- 100
- Repository forks
- 11
- 7-day trend score
- 0
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/evanbaker315/attentionos/linkedin-longform)