Threads post writer
Draft a single Threads post or a multi-post thread using a 2026 Threads hook formula (warm contrarian, data-point, build-in-public, quote-post, mini-list, relatable, listicle, story, curiosity-gap, how-I teardown), picked by goal (replies, reposts, likes, quotes). Respects the 500-char limit and one-hashtag cap, runs the humanizer pass, and publishes via Publora, which auto-splits long content into a thread. Not for auditing a draft (use threads-humanizer) or hook teardown (use threads-hook-extractor).From its SKILL.md
npx -y skills add sergebulaev/threads-skills --skill threads-post-writerAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 19 stars19 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
8.8 KB, as published. Nobody here has run it
Threads Post Writer
Ship a single Threads post, or a multi-post thread, using hook shapes that travel on Threads in 2026. Single-post formulas land a complete punch in the first line, because the feed truncates long posts with a "more" cut. Threads posts run warmer and more conversational than X.
When to use
- User says "write me a Threads post about X"
- User has a topic or a rough one-liner and wants a sharper hook
- User wants to turn notes into a multi-post thread
- User wants a quick draft that auto-publishes on approval
Formulas this skill uses
Single-post shapes (T1-T6):
| Code | Formula | Primary goal | Best for |
|---|---|---|---|
| T1 | Warm Contrarian | reposts | a sharp, defensible take framed as an invite |
| T2 | Data-Point Hook | reposts | one odd-precision number that reframes a thing |
| T3 | Build-in-Public Confession | replies | a real metric from your own work, ugly ones included |
| T4 | Quote-Post Add-Value | quotes | adding a layer to someone else's take (quote post) |
| T5 | Mini-List Post | reposts | a scannable list that fits in one post |
| T6 | Relatable Cold-Open | likes | a specific shared moment, no setup |
| T11 | First-Person Status Flex | replies | a flat status or milestone, then a counterintuitive pivot |
| T12 | Permission Reframe | quotes | "Most {group} DON'T NEED {thing}, they need {permission}" |
| T13 | Copy-Swap Micro-Lesson | reposts | one before/after swap: weak version, then strong version |
Thread shapes (T7-T10):
| Code | Formula | Primary goal | Best for |
|---|---|---|---|
| T7 | Listicle-Thread Promise | reposts | numbered teaching thread, one item per post |
| T8 | Story Thread | likes, quotes | a real narrative with a build and a turn |
| T9 | Curiosity-Gap Opener | replies | a surprising result, mechanism withheld |
| T10 | How-I Teardown | reposts | a repeatable process you ran yourself |
Full skeletons in ../../references/hook-formulas.md. T11-T13 were added from a
318-post corpus pull (July 2026): the status flex, permission reframe, and copy
swap topped the aspirational-warm Threads feed. They are strong hypotheses to
test, anchored on small-account examples, not yet follower-normalized.
Pick by goal first
| Goal | Reach for |
|---|---|
| Replies | T3, T9, T1, T11 |
| Reposts | T2, T5, T7, T10, T13 |
| Likes | T6, T8 |
| Quotes | T4, T8, T12 |
Pick the container first
| If the idea is... | Container | Formulas |
|---|---|---|
| One claim, one number, one moment | single post | T1, T2, T3, T6, T11, T12 |
| A scannable list, one line per item | single post | T5 |
| One before/after copy swap | single post | T13 |
| A list where each item needs teaching | thread | T7 |
| A narrative with a build and a turn | thread | T8 |
| A surprising result with a mechanism | thread | T9 |
| A repeatable process you ran yourself | thread | T10 |
A thread on a one-line idea reads as padding. A single post trying to teach five things should be a thread.
How Publora splits a thread
Pass the full thread as one content string. Publora auto-splits it:
- at paragraph breaks (
\n\n) first, - then at sentence endings,
- then at word boundaries as a fallback.
Unlike X, Threads does NOT add (1/N) markers by default. For full control of
where each post breaks, hand-split with a --- separator on its own line between
posts. The skill defaults to hand-splitting with --- so the user sees exactly
how the thread will land before approving.
If multi-post nesting is temporarily paused (Meta app reconnection), fall back to posting the opener as a single post and adding the rest as replies by hand. See
../../references/algorithm-heuristics.md.
Steps
Voice profile first (all drafts). If ../../references/voice-profile.md has filled: yes, load it and match the user's voice fingerprint, hard rules, and CTA/link style throughout. If it is not filled, mention once that threads-humanizer --mode profile can learn their voice from a few posts, then proceed with the generic voice rules.
- Gather inputs. Topic, angle, any rough draft, target audience (builders / founders / a niche), and the goal (replies / reposts / likes / quotes).
- Pick the container. Use the container table. One claim or moment is a single post; a list that needs teaching or a story with a build is a thread.
- Pick the formula. Use the goal table to shortlist, then suggest 2-3 that also fit the topic and let the user choose.
- Draft. Fill the skeleton in the user's voice. Respect the 2026 Threads
rules:
- First line carries the whole load (the feed truncates with "more").
- Under 500 chars per post (10,000 only if a text attachment is in play).
- One idea per post. Line breaks as beats.
- 0-1 hashtag at the end (Threads allows only one), 0-2 emoji, none on a serious take.
- No external link in post 1 (offer to put it in a reply).
- Warm and conversational, not a transplanted X dunk.
- For a thread: write post 1 first (the promise + open loop), front-load the strongest beat at position 1-2, target 4-7 posts for teaching/list threads, make each post stand alone, and close on the most quotable line plus one ask.
- Humanizer pass. Strip em dashes, AI vocab, rule-of-three, generic openers. Add a specific number or named entity where the claim allows it. Vary post length across a thread.
- Optional audit. Invoke
threads-humanizer --mode auditfor a pass-fail check. - Approval card. Show: formula used, full draft (thread shown post-by-post
with per-post char counts and total post count), suggested posting window,
primary goal. Confirm any
---breaks. - On approval. Call
lib.publish(kind="post"(single) orkind="thread",draft_text=<approved>, target_url="https://www.threads.com/", platforms=[<THREADS_PLATFORM_ID>], scheduled_time=<iso_or_None>, platform_settings=<{"threads": {"replyControl": "everyone"}} or None>). The wrapper handles Publora / manual / diy routing. If the content runs long, Publora auto-splits it into a connected thread.
Hard rules
Global voice rules: see root SKILL.md Voice rules. Additional skill-specific
rules:
- The first line must stop the scroll on its own. Rewrite any opener that needs line 2 to make sense.
- Keep each post inside 500 chars. If it does not fit, tighten it or escalate to a thread, never ship a truncated thought.
- One hashtag maximum. A second is rejected by Threads, not just bad style.
- One specific number where the claim allows it. 2.4x beats "way better".
- Do not hard-sell the user's product. One natural mention max.
- For a thread, post 1 must contain a real promise and an open loop. A bare "a thread:" label is not a hook.
Anti-patterns (skill will refuse)
- ALL CAPS first line for intensity. Carry intensity with word choice.
- Em dashes anywhere.
- "Unpopular opinion:" on a take that is actually popular.
- Padding a one-line idea into a thread.
- Rule-of-three lists without specifics.
- "leverage", "fundamentally", "game-changer", "deep dive".
- External link in post 1.
- A second hashtag.
- Engagement bait ("repost if you agree", "reply YES").
- A thread of identical-length posts (machine rhythm).
- Importing a cold, combative X voice.
Resources
../../references/hook-formulas.md- all 13 Threads formula skeletons (T1-T6 and T11-T13 single, T7-T10 thread)../../references/algorithm-heuristics.md- 2026 Threads ranking rules (signals, timing, limits)references/single-post-checklist.md- the per-post scrub and fit listreferences/thread-structure.md- per-position pacing and the closer playbook
Optional illustration
Offer a generated image when a visual would lift reach. Draft a prompt and call
lib.illustrate(prompt, kind="square"), pulling brand handle/color from Voice &
Brand Profile section 6 for a pixel-exact overlay. Show the returned url + cost,
then attach on publish via media_urls=[url]. Full workflow (incl. quote-cards):
../threads-humanizer/sub-skills/illustration.md. No Pixfaro key -> it drafts the prompt for you to generate manually.
Related skills
threads-humanizer- aggressive AI-tell scrubber, plus--mode auditfor reviewthreads-hook-extractor- reverse-engineer a hook from a post you admirethreads-content-planner- schedule posts and threads into a weekly cadence
What ships with it: 2 files
4.7 KB alongside SKILL.md
references/
- single-post-checklist.md1.8 KB
- thread-structure.md2.9 KB
Gives 0 of the 12 instructions most docs writing skills give
Counted across 1,951 of the 3,904 authors here whose files we hold, read 2026-09-06
- Use third-person for skill descriptionsin 54 of 1951, across 35 files
- Start descriptions with Use whenin 43 of 1951, across 29 files
- Run baseline scenarios before writing any skillin 40 of 1951, across 26 files
- Use active voicein 40 of 1951, across 36 files
- Map file responsibilities before defining tasksin 36 of 1951, across 29 files
- Use checkbox syntax for tracking stepsin 35 of 1951, across 27 files
- Ask one question at a timein 35 of 1951
- Offer execution options after saving the planin 33 of 1951, across 24 files
- Include complete code in every stepin 33 of 1951, across 27 files
- Design units with clear boundaries and interfacesin 31 of 1951, across 23 files
- Announce the skill usage at the startin 30 of 1951
- Verify agent compliance after adding the skillin 29 of 1951, across 17 files
Said here and by no other author read
- Use hook formulas based on engagement goals
- use one hashtag maximum
- Pick a container based on the idea complexity
- Keep each post under 500 characters
- Strip AI vocabulary and em dashes
- Vary post length throughout a thread
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once.