milad-ahmd/agent-skills / skills/social-media-manager
social-media-manager
FlaggedHead-of-social agent that picks which platforms a brand should focus on, sets KPIs, runs the weekly plan and community work, and does the hands-on work on LinkedIn, X (Twitter), Facebook Pages and Groups, Pinterest, Threads, Reddit, Telegram channels, WhatsApp Business and Google Business Profile, loading one platform guide at a time. Instagram, TikTok and YouTube work goes to their own skills. Use when the user asks for overall social media management, a social strategy, "manage my socials" or which platforms to use, or for work on one of those nine platforms, such as a LinkedIn headline or post, tweets or an X thread, a Facebook Page, Group or Messenger replies, Pin titles and boards, Threads posts, promoting on Reddit or finding subreddits, growing a Telegram channel, WhatsApp broadcasts, templates or chatbot flows, ranking on Google Maps or replying to Google reviews.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 5.9 KB |
| references/facebook.md | markdown | 4.7 KB |
| references/google-business-profile.md | markdown | 4.7 KB |
| references/linkedin.md | markdown | 4.4 KB |
| references/pinterest.md | markdown | 4.1 KB |
| references/reddit.md | markdown | 4.0 KB |
| references/telegram.md | markdown | 4.4 KB |
| references/threads.md | markdown | 3.0 KB |
| references/whatsapp-business.md | markdown | 4.9 KB |
| references/x-twitter.md | markdown | 4.3 KB |
references/linkedin.md
View on GitHub# LinkedIn
For a personal brand or a company page: profile rewrites, posts, PDF carousels, comment and DM strategy, newsletters and analytics reviews. B2B, hiring, job search and thought leadership.
## Platform facts
Platforms change limits. If a number matters for a deliverable, tell the user to check it in the composer.
| Item | Limit / note |
|---|---|
| Personal headline | 220 characters |
| About (personal) | 2,600 characters |
| Post | 3,000 characters. Only the first 2–3 lines show before "...see more" (fewer on mobile) |
| Comment | 1,250 characters |
| Connection note | 300 characters. Free accounts get a limited number of personalised notes per month |
| Company page tagline / About | 120 / 2,000 characters |
| Document post (carousel) | Upload a PDF; each page is a slide. Square or 4:5 portrait pages read best on mobile |
## How visibility works
- The feed first shows a post to part of the author's network, then widens it if people engage. Comments with substance count for more than likes, and replies from the author keep the thread alive.
- Posts that keep people on LinkedIn tend to do better. Many creators put external links in the first comment. This is widely reported, not confirmed by LinkedIn, and it changes, so test it on the user's own account.
- Hashtags matter much less than they used to. Use 0–3 relevant ones or none.
- Relevance to the poster's stated expertise matters: a consistent set of topics beats random viral attempts.
## Intake (save to `social/linkedin.md`)
- Goal: leads, hiring, a job, authority or a network.
- Target audience by role and industry.
- Topics they can speak on with authority, and proof: results, numbers, stories. Only what the user gives you.
- 3 past posts or messages, to match the voice.
- Personal profile, company page or both. Current followers and hours per week.
## Deliverables
**Profile rewrite**
- Headline: who you help, the outcome, the proof. Not just a job title. 3 options.
- About: hook line, story and credibility, what you do for whom, CTA with contact details. First person.
- Featured: best post, lead magnet, case study or product.
- Experience: outcome bullets using only numbers the user provides. Banner and custom URL.
- Company page: tagline, About, specialties, CTA button, showcase pages if there are distinct product lines.
**Content plan**: 3 pillars (expertise / story / opinion), 3–5 posts a week, a mix of text stories, PDF carousels, short native video, and polls only when the answer is useful. A newsletter only if the user can commit to the cadence.
**Post formula**
```text
Line 1: hook, under ~12 words, readable before "...see more"
Line 2: re-hook or tension
Body: short lines, whitespace, one idea, a specific story or data point
Takeaway
CTA: a question that invites a real answer
0–3 hashtags. External link in the first comment if reach matters (test it)
```
**Engagement system**: about 20 minutes a day of comments of two sentences or more on 5–10 posts by people in the target audience, plus replies to every comment on the user's posts, ideally in the first hour. Connection notes under 300 characters, personal, no pitch. DM sequence: value first, ask permission before pitching.
**Analytics review**: impressions, engagement rate, follower growth, profile views, viewer job titles and industries, top posts and why they worked, then 3 experiments for the next 2 weeks.
## Policy and ToS
- LinkedIn's User Agreement bans bots, scrapers and browser extensions that automate activity (auto-connect, auto-comment, auto-DM). Engagement pods break the Professional Community Policies on inauthentic behaviour. Both can get the account restricted.
- Never invent achievements, metrics, titles or client names.
- Mark sponsored or paid content as such.
## Tools and posting
- LinkedIn has a native post scheduler. Buffer, Hootsuite and similar tools post through LinkedIn's official API.
- Direct API posting needs an approved LinkedIn app (Community Management / Posts API). Don't assume the user has it.
- Draft everything; post or schedule only through tools the user has connected, after they confirm each post.
## Review checklist
- [ ] Hook works on its own in the first two lines
- [ ] One idea per post, in the user's voice
- [ ] Every number, name and result came from the user
- [ ] CTA asks a question people can actually answer
- [ ] No automation, pods or mass-connect advice
- [ ] User has approved before anything is posted
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |