substack
Skill18 starsChanged 4 months agoMITView on GitHub
Generate a weekly newsletter issue (or a short note) in the user's voice, ready to paste into Substack. Finds the week's signal, picks the angle, writes the draft to a local file, and gives clean publish steps. Use when the user says /substack, 'write the newsletter', 'draft this week's issue', 'substack this week', or 'write a newsletter note'.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- AceDataCloud/Skillstheir own version, first seen Oct 6, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- AceDataCloud/Skills★ 17same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository52 other files it gives its agents.
- ad-studioSkill
- brainstormingSkill
- build-siteSkill
- carousel-creatorSkill
- cold-outreachSkill
- content-engineSkill
- content-promptSkill
- daily-digestSkill
- daily-post-ideasSkill
- defuddleSkill
- digital-guide-builderSkill
- dissectSkill
- dm-writerSkill
- edu-siteSkill
- faceless-videoSkill
- first-customer-closerSkill
- frontend-designSkill
- higgsfield-generateSkill
- hooksSkill
- ig-channel-remixSkill
- infographicSkill
- launch-planSkill
- lead-magnetSkill
- linkedin-post-batchSkill
- linkedin-postSkill
- meeting-notesSkill
- money-model-auditSkill
- objection-handlerSkill
- offer-builderSkill
- one-lead-magnetSkill
- outlinesSkill
- personalizeSkill
- positioning-filterSkill
- pricing-calculatorSkill
- productize-serviceSkill
- proposal-generatorSkill
- saas-ad-studioSkill
- sanitize-projectSkill
- shortifySkill
- ui-ux-pro-maxSkill
- viral-format-minerSkill
- viral-remixSkill
- warm-outreachSkill
- web-design-guidelinesSkill
- yapSkill
- youtube-clipping-agentSkill
- yt-descriptionSkill
- yt-introSkill
- yt-searchSkill
- yt-thumbnailSkill
- yt-titlesSkill
- skill-nameSkill
The file
We split this skill into the rules it gives an agent
8 rules. About 0 are common, 7 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈7
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 19
If it doesn't exist, run the personalize skill first (or just tell me your niche, audience, and what you're building — I'll create it).
- Distinctive: line 24
If a field this skill needs is blank, ask one quick question and save the answer back to the profile.
- Distinctive: line 49
No em dashes, no filler, short sentences, line breaks.
- Distinctive: line 61
A) What the user worked on this week: Ask: "What did you build, ship, learn, or figure out this week?" (One line is enough.) If the user …
- Very specific: line 93
Write the full issue in the user's voice.
- Distinctive: line 133
Never hedges, never talks down.
- Distinctive: line 136
No em dashes. No "not X, but Y" constructions.
- Distinctive: line 141
Never use real client/member names — "a client who…", "someone I work with who…".
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Weekly Newsletter Generator
Finds the signal, picks the angle, writes the draft in the user's voice, and saves it as a clean Markdown file the user pastes into Substack. No external publishing scripts required.
Invoke with no args for the weekly article. For a short note: /substack note.
Personalization
This skill writes in your voice, about your week. Before running, load the brand profile at
~/.claude/brand-profile.md.
- If it doesn't exist, run the personalize skill first (or just tell me your niche, audience, and what you're building — I'll create it).
- Pull Tone, Words-to-avoid, Audience, Transformation, Proof, and Default-CTA from the profile. Every hook and the closing CTA must come from real facts in the profile — never invent results, numbers, or client wins.
- If a field this skill needs is blank, ask one quick question and save the answer back to the profile.
Requirements
- Publishing is manual by default. This skill writes a ready-to-paste Markdown draft to a local file; the user copies it into the Substack editor and hits publish. No API keys needed.
- Optional: if the user has their own Substack automation (a script, a Zapier/Make flow), they can wire the final file into it — but nothing here depends on it.
NOTE MODE (/substack note)
A short, punchy note (50-200 words) — one scene, one point.
- Pick a target length: randomly choose 50, 100, 150, or 200 words. State it before drafting.
- Find the moment: pull one real moment from the profile's Proof/Transformation, or ask the user for one thing that happened this week worth writing down. Notes are grounded in a real moment, never an abstract observation ("most people do X" is filler, not a hook).
- Draft it in the user's voice:
- Hook: lead with the transformation or the moment in concrete terms (real numbers where the user has them). Make the reader want it.
- Body: diary-like, personal, plain language. What actually happened. No quote-card aphorisms.
- Close: hand the reader something concrete — how this benefits them and what they could do differently. Never assume their situation; state what's true for the user and let them draw the parallel.
- No em dashes, no filler, short sentences, line breaks. Reads in 60 seconds.
- Save to
~/newsletter/notes/<YYYY-MM-DD>-<slug>.txtand show it to the user to paste into Substack Notes.
WEEKLY ARTICLE MODE (default)
Step 1 — Gather this week's context (run in parallel)
Fire these together:
A) What the user worked on this week. Ask: "What did you build, ship, learn, or figure out this week?" (One line is enough.) If the user keeps a work log or project notes, offer to read a path they give you. This becomes the "What I Built / Learned This Week" anchor — the realest, most specific part.
B) What's happening in the user's niche this week — 2-3 WebSearches scoped to their field, e.g.:
[user's niche] new tool OR release OR update after:[date 7 days ago][user's niche] news this week site:techcrunch.com OR site:theverge.com- one search on a specific topic the user is watching
C) The user's own recent content (optional). If they want to embed a video/post, ask for the URL.
Step 2 — Pick topic + built anchor
Choose the "What I Built This Week" anchor by this priority:
- A real build/ship from this week (best — real, specific, theirs).
- An idea or insight that clicked this week (strong).
- A niche news drop to react to (fallback — commentary, never a fabricated personal build).
Then pick a topic angle that connects naturally to that anchor and hasn't run in recent issues. Good evergreen angles (adapt to the user's niche):
- Attention that doesn't convert (engagement vs. income)
- The build-before-launch trap (building instead of selling)
- The sequencing mistake (tools before offer, complexity before proof)
- Permission to ship imperfect
- Selling the outcome, not the tool
- Niche clarity — picking a lane
The built anchor illustrates the topic; the topic reframes what the build revealed. They should feel like one idea. State the chosen topic + anchor + why they connect before drafting.
Step 3 — Write the draft
Write the full issue in the user's voice. Follow the ARTICLE STRUCTURE below. Save to
~/newsletter/<YYYY-MM-DD>-<topic-slug>.md. Print the path and the first ~300 characters so the user
can sanity-check the voice.
Step 4 — Report back
Issue drafted.
Topic: [topic angle]
Anchor: [built anchor — one line]
Hook: [first line]
Proof: [any real win/number used]
File: [path to .md]
To publish: open Substack → New post → paste the Markdown → add a cover image (optional) → publish.ARTICLE STRUCTURE REFERENCE
Section order (keep it):
- Opener: Hook → open loop → "this week:" preview bullets → loop closure (what it means → why it matters → the shift)
- New This Week (embed a video/post here if the user gave one)
- One Thing Worth Stealing — closes with WHY + the dream outcome (specific: first client, replaced income, etc. — never vague)
- What I Built / Learned This Week — story arc: problem → what I did → the teaching moment. A blockquote with a real number/stat if one exists, and it must NOT echo the opener. This must be a DIFFERENT build from the one embedded in "New This Week."
- A Win (the user's own or a client's, with permission and no real names — "a client who…") — the WHY moment before the result; the transformation is what they now KNOW, not just what they got
- Pre-CTA question
- CTA — the user's Default-CTA from the profile
Voice rules
- Opinion-first — lead with the point, not the setup.
- Short sentences. Specifics over generalities. Real numbers only (the user's actual figures).
- Never hedges, never talks down. Treats the reader as capable.
- Sounds like a diary entry or a text from someone who knows what they're doing — not a marketing email, not a quote card.
- No em dashes. No "not X, but Y" constructions. No preachy closing lines — end on something plain.
- Be concrete: name the actual actions a person takes (ship the feature, message a stranger, name a price, sit with a no), not abstractions like "building feels like progress." If a sentence could apply to almost anything, it's too vague — replace it with the specific thing. This is the most common failure in generated drafts; check every paragraph.
- Never use real client/member names — "a client who…", "someone I work with who…". Establish who they are before any pronoun.
- Respect the profile's Words-to-avoid.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

