/fb-humanizer

name: fb-humanizer
description: "Remove the AI tells readers react to in a Facebook Page post: 2026 vocabulary by density, reveal bridges, staccato stacks, stacked triads, performed sincerity, \"We are thrilled to announce\" auto-pilot; caps em dashes. Includes --mode audit (under-80 sweet spot, hook, engagement bait, hashtag and emoji limits) and --mode profile. Not for beating AI detectors (no edit reliably does). Not for writing from scratch (use fb-post-writer). Keywords: humanize, de-AI my post, audit before posting."

Facebook Page Humanizer V3

Rewrites any Facebook Page post to remove the AI tells that human readers notice, and audits a finished draft against the 2026 Facebook ranking checklist. Based on Wikipedia's "Signs of AI writing" taxonomy, the 2025-2026 stylometry literature, and Facebook-Page-specific patterns (the under-80 sweet spot, the "See more" fold, the "We are thrilled to announce" corporate tell, and the meaningful-interactions model). **V3 (2026-09):** recalibrated on 2026 evidence. Vocabulary is scored by density, em dashes are capped instead of banned, forced rhythm is now a tell instead of a fix, and there is an over-correction guard. Facebook has no corpus of its own yet, so the calibration follows the LinkedIn one (the closest long-and-short mixed feed).

**What this skill does not do:** it does not make text "pass" GPTZero, Pangram, Turnitin or Originality. Those are trained classifiers keyed on the instruction-tuning style signature; prompt-style "sound like a real person" rewrites are caught 92-95% of the time, and light mechanical rewriting raises detectability. On post-length text (under 300 words) detector scores are noise. The real value is elsewhere: expert human readers cite vocabulary (53%) and sentence structure (36%) as what gives AI text away, and a Page post that reads as a bot earns neither the share nor the comment Facebook's meaningful-interactions model ranks on. This skill removes what those readers react to.

What changed in V3

Evidence tier in brackets: [strong] = replicated across 2+ independent 2025-2026 studies or our own corpus on sibling platforms; [vendor] = single platform or vendor dataset; [weak] = one study or expert-panel report.

• **Vocabulary moved from a delete-list to density scoring.** The 2023-24 words

(delve, tapestry, realm, journey) are decaying as humans avoid them [strong]. The durable 2026 markers are common words (significant, crucial, notably, comprehensive, insights, robust, leverage, foster, landscape, nuanced, streamline, elevate) plus grammar: nominalisations and "-ing" clause openers at 5.3x the human rate [strong]. AI vocabulary is the one marker consistently reach-negative on the sibling platforms we measured [strong]. One marker in a paragraph is not a verdict. Three is.

• **Em dash is no longer a tell.** GPT-5.4 emits 1.43 per 1,000 words, below

the 3.23 human baseline; 23-29% of human posts and captions on sibling platforms use one [strong]. Zero em dashes is now its own tell (the writer is trying to look human). New rule: cap at about 1 per 100 words (so 0-1 in a short post, 1-2 in a story post), replace only the excess with a comma, colon, parentheses or a rewrite. Never a period.

• **Forced burstiness is the #1 2026 tell, not the fix.** Mechanical

long/short alternation is a learnable humanizer fingerprint [weak], and on the platforms we measured sentence-length variance is not an engagement lever in either direction (LinkedIn within-creator: null; Threads: uniform wins) [strong]. "Short. Punchy. Done.", "No X. No Y. Just Z.", one-word lines for drama and "The result?" reveals are the current top tells. Pass 2 is now RHYTHM, not BREAK: an anti-uniformity guard only, never manufactured variance. A short Page post has nothing for Pass 2 to touch.

• **Rule of three is still a tell, at density.** Tricolon runs at 2x

expert-human rate across 2026 frontier models [strong]. Stacked, perfectly parallel triads and 3+ per post get scrubbed. One natural triple with concrete items stays (21-39% of top human posts have one).

• **Fingerprint injection was half wrong.** Named entities and concreteness are

supported [strong]; an odd-precision number with a referent in line 1 is the strongest opener [vendor]. Bare numbers are not a discriminator, and inserted hedges and confessions backfire: performed hesitancy is 2x more common in LLM text, and sincerity announcements ("let me be honest", "we'll be real with you") are a named 2026 tell [strong]. Pass 3 asks for a flat, dated, uncomfortable fact instead.

• **Over-correction guard.** Humanizer output has its own fingerprint [weak].

Pass 4 checks whether Passes 1-3 introduced the very patterns they were meant to remove. Edits are proportional to real problems. When in doubt, leave it.

When to use

• Before publishing any AI-drafted Page post (rewrite mode)
• Pre-publish review of a finished draft (audit mode, see `sub-skills/post-audit.md`)
• When a draft feels corporate or off and you cannot pinpoint why

Input

Any text: a short Page post, a longer story post, or a comment reply draft. Optional: target voice samples (the Page's past posts).

Output

• Rewritten text with AI tells removed
• A diff showing what changed and why
• Char count, with a flag when a post crosses the 80-char sweet spot
• Per-paragraph tell density (markers per paragraph; 3+ triggered a rewrite)
• Reader-read confidence: "reads human", "mixed", "reads AI" (a reader-tell

estimate, not a detector score)

Modes

# Default: scrub AI tells (forensic + strict) and fix Facebook-format issues
fb-humanizer <text>

# Forensic only - minimum touch, just kill model leakage
fb-humanizer --mode forensic <text>

# Audit - detection-only pass-fail review, no rewrite
# Runs the 2026 Facebook checklist: under-80 sweet spot, first-line hook,
# engagement bait, hashtag/emoji limits, link-post reach warning, goal cla