agentsclimarketplace
Sign in

Jab hook

Skill coreyhaines31/makerskills/skills/jab-hook

Gary Vaynerchuk's jab-jab-jab-right-hook framework applied to a personal portfolio rotation on X and LinkedIn. Jabs = build-in-public + educational (value). Hooks = promo (the ask). Each property in the user's configured portfolio (see `~/.config/makerskills/jab-hook/properties.yaml`) gets a hook at least once every ~3 weeks; jabs fill the rest. Drafts go into the user's Typefully workspace via MCP. Modes — plan (7-day plan), pick-next (single post), audit (coverage report), draft (specific post). Triggers on "/jab-hook," "what should I post," "plan my socials," "next promo," "next jab," "next hook," "social rotation," "promote [property]," "BIP post," "audit my socials," "what haven't I posted about."From its SKILL.md

Install
npx -y skills add coreyhaines31/makerskills --skill jab-hook

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • reads credentialsReads from 1 credential source: `${MAKERSKILLS_CONFIG:-$HOME/.config/makerskills}/jab-hook/typefully.yaml`.
  • runs commandsInstructs the agent to run 3 commands, including `mcp__typefully__typefully_list_social_sets` and 2 more.

SKILL.md

7.9 KB, ~1.8k tokens by cl100k_base, as published. Nobody here has run it

/jab-hook — Jab-jab-jab-right-hook for a personal portfolio

Gary Vaynerchuk's framework applied to your configured portfolio: keep the feed mostly jabs (value — BIP + educational) so the hooks (promo) earn attention when they land. Drafts to your personal Typefully workspace.

One-time setup: see references/properties.md and references/typefully-config.md to configure your real portfolio + workspace. Personal config lives in ${MAKERSKILLS_CONFIG:-$HOME/.config/makerskills}/jab-hook/.

Mental model

  • N rotation slots (configured in ~/.config/makerskills/jab-hook/properties.yaml): one per property you want to promote on rotation
  • ~2 promo posts/week → each property cycles every ~3 weeks at equal weight
  • Non-promo days: 50/50 build-in-public vs educational
  • Target weekly rhythm: ~2 promo + ~2–3 BIP + ~2–3 educational
  • Platforms: X + LinkedIn via your personal Typefully workspace

Step 1 — Pick the mode

Detect from the trigger:

User saysMode
"plan my socials," "next week's posts"plan
"what should I post," "next promo," "pick next"pick-next
"audit my socials," "what's overdue," "what haven't I posted about"audit
"draft a promo for X," "BIP post about Y," "educational about Z"draft

If ambiguous, confirm.

Step 2 — Load context

  1. Read references/properties.md — the 6 rotation slots and angle ideas
  2. Read references/voice.md — the user's voice rules per platform
  3. Read references/content-types.md — templates for promo / BIP / educational
  4. Pull recent posts from your personal Typefully workspace:
    • First run: call mcp__typefully__typefully_list_social_sets and ask which social set is your personal (X + LinkedIn). Save the ID to ${MAKERSKILLS_CONFIG:-$HOME/.config/makerskills}/jab-hook/typefully.yaml for future runs (schema in references/typefully-config.example.yaml — never save it inside the skill folder; upgrades wipe it).
    • Call mcp__typefully__typefully_list_drafts filtered to the last 30 days
    • Classify each as promo / BIP / educational by content
  5. Compute days since last promo for each of the 6 slots

Step 3 — Mode logic

plan (7-day plan)

  • Slot the 1–2 most overdue properties as promo posts
  • Fill remaining 5 days with alternating BIP and educational (50/50)
  • Output as a table: Day | Type | Property/Topic | Hook | Draft summary
  • Don't stack two promo on consecutive days
  • After the user approves the plan, offer to draft each one in sequence

pick-next

  • Identify the most overdue property
  • If most overdue < 14 days, suggest a BIP or educational instead (whichever is less recent)
  • Draft 1 post; ask before syndicating

audit

  • Report: days since last promo per property, ordered by most overdue
  • Flag any property >21 days as overdue
  • Suggest the next 1–2 moves

draft

  • Skip rotation logic — draft the requested post
  • Use the right content-type template

Step 4 — Draft

Follow references/voice.md and references/content-types.md. Default: single post, one per platform (X version + LinkedIn version). Thread only if the second post earns its place.

Link placement (load-bearing): never put a URL in the body. Default is no link at all — let curiosity drive the click. When a link is needed, draft it as a first comment for LinkedIn (Typefully supports auto-comment) or a reply for X. The body always stands alone.

Optional: inspiration scan

For promo and educational posts, optionally scan inspiration accounts (references/inspiration.md) before drafting:

  1. Pick 1–2 accounts whose audience overlaps with the property
  2. Pull recent posts (agent-browser if it works on LinkedIn; otherwise ask the user to paste examples or fall back to their X/newsletter)
  3. Extract structural patterns: hook openers, post length, line-break rhythm, CTA styles
  4. Apply ONE pattern to the user's voice — never mimic phrasing
  5. Log new patterns in ${MAKERSKILLS_CONFIG:-$HOME/.config/makerskills}/jab-hook/patterns.local.md (create if missing) so they accumulate — references/patterns.md documents the starter patterns and stays read-only

Skip inspiration scan when the user says "just draft it" or when iterating on an existing post.

Step 5 — Syndicate to Typefully

Ask: "Push to Typefully now? (X + LinkedIn, your personal workspace)"

If yes:

  1. Read social set ID from ${MAKERSKILLS_CONFIG:-$HOME/.config/makerskills}/jab-hook/typefully.yaml (setup notes in references/typefully-config.md)
  2. Call mcp__typefully__typefully_create_draft once for X, once for LinkedIn (or once with both platforms if the social set spans both)
  3. If the draft has a first-comment link, configure Typefully to add it automatically (LinkedIn auto-comment / X reply). If the MCP doesn't support auto-comment for one of the platforms, surface the comment text so the user can paste it after publishing.
  4. Default to draft state (not scheduled) — the user reviews in Typefully UI before sending
  5. Return the Typefully draft URLs

Composes with

  • paste — clean output for pasting into Typefully / LinkedIn / X manually if the MCP path doesn't fit
  • social-fetch — pull inspiration-account posts for structural analysis (replaces inline agent-browser)
  • second-brain — the Content Ideas wiki page hoppers hooks, frameworks, stories; /jab-hook drafts pull candidates from there
  • deep-research — when a promo needs a stat or citation the draft doesn't have yet
  • marketingskills:social — generic social frameworks, useful when teaching strategy vs shipping your own posts
  • marketingskills:copywriting — for hook / headline ideation when stuck

Notes on quality

  • 1 promo per week per property is the floor, not the ceiling. Rotation compounds when it's predictable — 4 properties × 1 promo/week = 4 promo weeks/month. Educational + BIP + community fill the rest.
  • Never mimic phrasing from inspiration accounts. Extract structure (hook opener type, line-break rhythm, CTA style) and apply to your own voice. Voice is the moat; copying phrasing destroys it.
  • Links belong in first comments, not the body for LinkedIn + X. Documented in ~/.claude/memory/feedback_social_link_placement.md. Body posts with inline URLs get algorithm-suppressed on both platforms.
  • Draft state, not scheduled state. Every Typefully push defaults to draft — human review in the Typefully UI before publish is non-negotiable. Automated scheduling of unreviewed drafts has repeatedly produced posts the user regrets.
  • Reader-perspective framing. "You'll know X" not "we'll teach you X." "What you're guaranteed" not "what we'd guarantee."
  • Voice.local.md overrides the shipped voice.md. Personal voice rules never live in the public repo — always in ${MAKERSKILLS_CONFIG:-$HOME/.config/makerskills}/jab-hook/voice.local.md.
  • Cadence cap: 2 posts/day per platform, per account. More than that degrades reach + reads as spam.

Memory references

  • ~/.claude/memory/feedback_promo_voice.md (if present) — conviction-coded CTAs + reader-perspective framing
  • ~/.claude/memory/feedback_social_cadence.md (if present) — cadence cap (e.g., 2 posts/day per platform)

What ships with it: 9 files

20.5 KB alongside SKILL.md

references/

Gives 0 of the 12 instructions most skill authoring skills give in ~1.8k tokens

Counted across 408 of the 422 authors here whose files we hold, read 2026-09-06

  • Keep SKILL.md under 500 linesin 118 of 408, across 104 files
  • Write skill instructions in imperative formin 44 of 408, across 37 files
  • Move detailed reference material into references filesin 41 of 408, across 34 files
  • Run init_skill.py to initialize new skillsin 40 of 408, across 33 files
  • Draft assertions while test runs are in progressin 32 of 408, across 25 files
  • Keep description under 1024 charactersin 31 of 408, across 18 files
  • Run package_skill.py once the skill is completein 29 of 408, across 23 files
  • Spawn with-skill and baseline runs in the same turnin 28 of 408, across 21 files
  • Test added scripts by actually running themin 27 of 408, across 21 files
  • Follow the creation steps in orderin 24 of 408, across 17 files
  • Generate the eval viewer before evaluating outputs yourselfin 24 of 408, across 17 files
  • Write the description in third personin 24 of 408, across 18 files

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.