short-form-social

name	short-form-social
description	>-

Short-Form Social — Hook, Body, Channel-Fit

Purpose

This skill turns an idea into a platform-native short-form post — LinkedIn, X single, or X thread — that a technical/founder audience will actually read and pass on. It owns the part of growth content that lives or dies on the first line and on respecting the platform's native form.

It exists because the two failure modes are predictable: a strong idea wrapped in a hook that needs setup before it lands (so no one reads past line one), and a post cross-pasted across platforms without rewrite (so it reads as native to none). This skill fixes both — hook first, then a body shaped to the specific surface.

It composes with the rest of the content suite: voice-print gives the posts a voice; content-craft owns long-form (an essay is not a thread); newsletter-growth owns owned-channel email; ai-slop-audit de-slops the draft (social is where AI tells are most visible and most punished).

---

Inputs

Input	Required?	Description
The raw material	Required	The idea, observation, result, or longer piece to adapt. Specific beats abstract — a real number, a real scar, a real before/after.
Platform(s)	Required	LinkedIn / X single / X thread. Each has different native form; "post this everywhere" without per-platform rewrite is the main failure.
Goal	Recommended	Reach / signups / authority / replies. Shapes the CTA (or its absence).
VOICE-PRINT.md	Recommended	From voice-print. Generic social voice is invisible; the founder's voice is the differentiator.
ICP	Recommended	Who you're trying to reach — calibrates the hook's specificity and the platform choice.

BLOCK rule

IF the raw material is vague ("write something about AI") with no specific
observation, number, result, or story:
  BLOCK. Return: "Short-form that travels is built on one specific thing — a
  number, a scar, a before/after, a contrarian observation you can defend. What's
  the specific thing? 'Post about X' without a specific anchor produces generic
  content the audience scrolls past."

---

Decision Logic

Step 1 — Pick the platform and its native form

Don't write then cross-post; choose the surface, write to it. Full per-platform rules in references/hooks-and-formats.md; these defer to growth/DOMAIN.md distribution norms and content-review platform-format discipline (single source of truth — don't restate them, apply them). Summary:

Surface	Native form	The one rule people break
LinkedIn	One insight; professional opener; 150–300 words; optional single CTA	Manufactured-curiosity opener ("You won't believe…")
X single	Hook stands alone; one claim; within char limit	Hook that needs a second tweet to make sense
X thread	Tweet 1 readable alone; one idea per tweet; last = takeaway	"Here are my N takeaways 🧵" opener; thread numbering; link in body (put it in the first reply)

Step 2 — Build the hook (the first line does 80% of the work)

The hook is the highest-leverage atom. The mechanism: the first ~8 words must carry a complete tension, claim, or fact — the reader decides to continue or scroll there. Hook patterns (detail + examples in references/hooks-and-formats.md):

Pattern	Shape	Best for
Specific number	A surprising concrete metric, first words	results, before/after
Counterintuitive claim	A defensible claim that cuts against the default belief	opinion, authority
Named pain	The reader's exact situation, in their words	reach within an ICP
Before/after	The gap between two states	migration / improvement stories

The hook gate: does the first line stand alone and carry tension without setup? If it needs the second line to make sense, it's not a hook yet. (This is exactly content-review's X/LinkedIn opener check — a draft failing it should not ship.)

Step 3 — Shape the body to the surface

• LinkedIn: observation/data opener → one insight developed → optional single specific CTA. Opinion is grounded in evidence/experience before it's asserted.
• X single: the whole thing is the hook + one payload. No warm-up, no thread-bait.
• X thread: tweet 1 is a standalone hook; each subsequent tweet carries one fact/observation/step; the last tweet is the takeaway or a soft CTA; the link goes in the first reply (body links get throttled). No "here are N takeaways" framing — open with the sharpest single claim instead.

Step 4 — Build-in-public posture

For founders, the content that travels is the work shown honestly, not the announcement. Share the scar and the number ("we shipped 40 alerts; engineers muted all of them — here's what we changed"), not the press release ("excited to announce our alerting feature"). Honesty and specificity are the reach mechanism here; polish without substance reads as marketing and gets scrolled.

Step 5 — Repurpose: write one story, adjust the frame

When adapting a longer piece (essay, newsletter issue, migration story), don't write a separate story per channel — write one story and adjust the frame (cross-ref the MLG principle in growth/DOMAIN.md). The essay's core observation becomes the thread's spine; the LinkedIn version foregrounds the practitioner angle; the X version sharpens to the single hardest claim. Same truth, surface-shaped.

Step 6 — Voice + de-slop pass

Run the draft against the VOICE-PRINT.md and through ai-slop-audit. Social is where AI tells are most visible — emoji-bullet scaffolding, "here are N takeaways," uniform rhythm, hype — and where the technical audience punishes them hardest.

---

Outputs

## Short-Form Social — [topic]

**Platform(s):** [LinkedIn / X single / X thread]   **Goal:** [reach/signups/authority]
**Voice print:** [used / none]

### Hook options
[2–3 hook variants, each a standalone first line, labeled by pattern. Mark the recommended one.]

### Post / thread
[Platform-native body. For a thread: numbered working draft (numbering removed before posting), one idea per tweet, tweet 1 standalone, last = takeaway.]

### Distribution note
[Timing; first-comment link (X thread / LinkedIn); any repurpose source.]

### Checks
[Hook gate pass? platform format pass? voice + ai-slop-audit pass?]

Cross-review: the draft goes through content-review (platform-format discipline + brand voice) and ai-slop-audit before posting.

---

Cross-Review Triggers

Before posting (customer-facing copy):

• content-review — platform format checks + brand voice
• ai-slop-audit — when AI-assisted (social tells are high-visibility)

---

Example Usage

Example 1 — Idea → X thread

Input: "Our monitoring philosophy: an alert that doesn't tell you what to do is noise." Platform: X thread. Output: Hook options (counterintuitive-claim recommended: "We deleted 39 of our 40 alerts. Incident response got faster."), thread with one idea per tweet, tweet 1 standalone, link in first reply, no "N takeaways" framing. De-slop + voice pass noted.

Example 2 — Essay → LinkedIn post

Input: A 2,000-word newsletter issue; wants a LinkedIn version. Platform: LinkedIn. Output: Same story, reframed for LinkedIn — practitioner observation opener, one insight, 200 words, single specific CTA (link in first comment). Not the essay truncated; the essay's spine re-shaped.

Example 3 — Vague brief

Input: "Write a post about why our product is great." Output: BLOCK — asks for the specific anchor (number, scar, before/after, defensible claim). Won't produce generic promotional content.

---

Validation Criteria

Output passes if:

• Platform chosen first; body is native to it (not cross-posted)
• Hook options provided; each stands alone and passes the hook gate (tension in first line, no setup)
• X threads: tweet 1 standalone, one idea per tweet, no "N takeaways" opener, link in first reply
• Build-in-public posture where relevant (scar + number, not announcement)
• Repurposing reshapes one story per surface (doesn't write separate stories or truncate)
• Voice + ai-slop-audit pass referenced

Output fails if:

• Same text proposed for multiple platforms without rewrite
• Hook needs a second line to make sense
• Emoji-bullet scaffolding / engagement bait / "N takeaways" thread
• Generic promotional post produced from a vague brief instead of blocking

---

Benchmarks

Metric	Benchmark	Source
LinkedIn thought-leadership engagement rate	1–5%; first 60 min determines reach	growth/DOMAIN (2025–2026)
X thread tweet-1 readable-alone	required for reach	growth-platform-execution-norms
HN/Show HN qualifying threshold (if used)	>20 upvotes to front page; strong first comment	growth/DOMAIN

---

References & Sources

Tier-2 (operator norms — adapted):

• growth-platform-execution-norms (growth-skills v1.0) — per-platform native rules
• Hook-pattern logic generalized (persona-stripped) from the author's claude.ai viral-hooks / short-form-video

Cross-references (this repo):

• references/hooks-and-formats.md — hook patterns + examples + full per-platform format rules
• growth/DOMAIN.md — distribution norms (SSOT for platform rules); MLG "write one story, adjust the frame"
• pmm/content-review — platform-format discipline (SSOT) + brand voice gate
• pmm/voice-print + pmm/ai-slop-audit — voice target + de-slop
• foundations/shared-skills/content-craft (long-form, distinct) · growth/newsletter-growth (owned email, distinct)

Output metadata:

---
Skill: short-form-social v1.0.0
Platform: [LinkedIn/X-single/X-thread]
Voice print: [used/none]
Generated: [date]
---