Skip to content

sergebulaev/x-skills / .codex-marketplace/x-skills/skills/x-humanizer

x-humanizer

Verified

Remove the AI tells human readers react to in a tweet or thread: 2026 vocabulary by density, reveal bridges, staccato stacks, stacked triads, performed sincerity. Also --mode audit (280-char fit and format), --mode profile (learns how you write) and --mode interview (fills your Story Bank: the numbers, turning points and positions posts are made of). Not for writing from scratch (use x-post-writer). Keywords: humanize, de-AI tweet, AI slop, interview me, audit before posting.

references/examples.md

View on GitHub
# Worked Humanizer Examples

## Example 1 - Single tweet, AI vocab + em dash

**Before:**
> In today's fast-paced world, founders must leverage AI tools to fundamentally
> streamline their workflow — it's a game-changer.

Tells: the tweet scores 4 markers ("in today's fast-paced world", leverage,
fundamentally, streamline) plus "game-changer", so the whole tweet is
rewritten rather than word-swapped; zero specifics; no first line that lands
alone. The single em dash was under the cap (one per tweet) and was not a tell
on its own; it went only because the sentence around it was rewritten. It was
not replaced with a period.

**After:**
> Most founders use AI to write faster.
>
> The ones who win use it to decide faster. Same tools, different job.

Why it is better: a complete one-liner contrarian (X1) take, lands in the first
line, no marker cluster, one clear idea, reads in a glance. Pass 2 left the
rhythm alone (single tweet: uniform rhythm is fine). Pass 4: nothing was
inserted, no hedge, no "not gonna lie" frame.

## Example 2 - Tweet over 280 from emoji

**Before** (looks like 276 display chars, but 4 emoji = 8 chars, so 284):
> shipping season is here 🚀🚀 we cut our deploy time from 40 min to 6 min and
> honestly it changed how the whole team works 🔥🔥 here is the one change that
> did most of it and why it matters for you

Tells: over 280 once emoji are counted as 2 each, emoji spam, two ideas, weak
trailing hook.

**After** (single tweet, under 280):
> we cut deploy time from 40 min to 6 min.
>
> one change did most of it: we stopped rebuilding the whole image on every
> push. here is how.

Or escalate to a thread if the "how" needs real steps (use x-thread-builder).

## Example 3 - Thread opener that closes its own loop

**Before (tweet 1):**
> Here are 5 ways to grow on X: post consistently, engage with others, write
> threads, use good hooks, and be authentic.

Tells: the whole payoff is in tweet 1, so no one expands; vague items; rule of
five with no specifics.

**After (tweet 1):**
> I grew from 200 to 11,000 followers on X in 7 months.
>
> 5 things did most of it. Number 3 is the one nobody talks about.

Now the loop is open (what is number 3?), tweet 1 promises a real result, and
the body has room to teach each item with a concrete example.

## Example 4 - Negative parallelism

**Before:**
> It's not about posting more. It's about posting better.

**After:**
> Posting more did nothing for me. Posting one sharp thing a day did.

Direct claim, first-person, a specific contrast instead of the AI "not X, it's
Y" rhythm.

## What to preserve

- The user's real claim and numbers.
- Lowercase-casual voice if that is their register.
- `..` soft pauses.
- Their actual story. Never invent a detail to make a tweet land.
rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first.Stars: 194kUpdated Claude CodeCodexVerifiedInstall