sergebulaev/linkedin-skills / .codex-marketplace/linkedin-skills/skills/linkedin-humanizer
linkedin-humanizer
VerifiedRemove AI tells from LinkedIn posts/comments: 2026 vocabulary density, reveal bridges, staccato fragments, stacked triads, performed sincerity. Tiered rewriter plus `--mode audit` LinkedIn post auditor and `--mode profile` voice builder. Not for drafting from scratch (use linkedin-post-writer) or beating AI detectors. Keywords: humanize, de-AI, post audit, post auditor, audit before posting.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 16.4 KB |
| references/audit-ai-tells.md | markdown | 11.6 KB |
| references/audit-checklist.md | markdown | 4.1 KB |
| references/audit-examples.md | markdown | 452 B |
| references/detector-list.md | markdown | 6.1 KB |
| references/emoji-patterns.md | markdown | 3.8 KB |
| references/examples.md | markdown | 1.4 KB |
| references/rules-explainer.md | markdown | 14.8 KB |
| references/scrub-rules.md | markdown | 27.0 KB |
| references/tier-rationale.md | markdown | 17.3 KB |
| references/v3-changes.md | markdown | 3.2 KB |
| references/voice-fingerprint.md | markdown | 5.3 KB |
| scripts/detectors.env.example | other | 428 B |
| scripts/test_detectors.pyexecutable | script | 13.0 KB |
| sub-skills/detector-tester.md | markdown | 6.0 KB |
| sub-skills/emoji-detector.md | markdown | 3.9 KB |
| sub-skills/illustration.md | markdown | 7.4 KB |
| sub-skills/post-audit.md | markdown | 4.9 KB |
| sub-skills/rules-explainer.md | markdown | 5.2 KB |
| sub-skills/voice-profile.md | markdown | 3.0 KB |
sub-skills/emoji-detector.md
View on GitHub# LinkedIn Emoji Detector
Flags AI-pattern emoji usage in LinkedIn drafts before they ship. Built on MagicPost frequency data (Feb 2026) showing lightbulb, rocket, sparkles, and recycling emojis appear 2-3x more often in AI-generated LinkedIn posts than in human-authored ones.
Pattern data from MagicPost LinkedIn analysis (Feb 2026, 220 engagement viral post).
## When to use
- Before publishing any AI-drafted post or comment
- As a pre-pass before `linkedin-humanizer` (catches a tell humanizer doesn't fix)
- When the audit pass flags "feels AI" without a specific reason
- When auditing a backlog of scheduled posts for AI signature emojis
## Input
Any LinkedIn text (post, comment, reply, DM). Optional: mode flag (`--strict`, `--lenient`, `--score`).
## Output
- AI-emoji density score (0-100, higher = more AI-like)
- List of flagged emojis with frequency vs. AI baseline
- Suggested human-pattern alternatives (or removal recommendation)
- Verdict: "clean", "borderline", "AI-likely"
## The three modes
### Mode 1 — SCAN (default)
Walks the text, extracts every emoji, looks up each one in the frequency table at `../references/emoji-patterns.md`, and returns a per-emoji report.
For each detected AI-pattern emoji:
- Show the emoji
- Show its AI-correlation frequency (e.g. lightbulb = 2.57%)
- Show count in the draft
- Suggest a human-pattern alternative or recommend deletion
### Mode 2 — SCORE (`--score`)
Returns a single number (0-100). No rewrite, no suggestions.
Formula:
- Each AI-pattern emoji contributes `frequency_pct * count * 10`
- Cap at 100
- Bonus +20 if 3+ distinct AI-pattern emojis present (cluster signal)
- Bonus +15 if a single emoji appears 2+ times (repetition tell)
Verdicts:
- 0-20: clean
- 21-50: borderline
- 51-100: AI-likely
### Mode 3 — SUGGEST
Returns a rewritten emoji set. Keeps the post's emoji count roughly the same but swaps AI-pattern emojis for less-correlated alternatives, or recommends removal where no clean substitute exists.
## Strict vs. lenient
### `--strict`
Flag any AI-pattern emoji at all. One lightbulb = AI-likely verdict. Use when content goes out under a personal brand that has never used emoji-heavy AI patterns.
### `--lenient`
Flag only if 3+ AI-pattern emojis appear in the same post, OR if a single AI-pattern emoji repeats 2+ times. Use for accounts where occasional emoji is fine and only the cluster pattern matters.
Default mode is between the two: flag if score >20.
## Non-negotiable rules
- Never auto-rewrite without showing the diff first
- Preserve emoji count parity if user wants emojis (don't strip and leave nothing)
- Capitalize all names in suggested replacements (HubSpot, Claude, etc.)
- If the post has zero emojis, skip the skill entirely and return "n/a"
- Don't fabricate "human" emoji frequencies — the human-pattern alternatives in `../references/emoji-patterns.md` are emojis with sub-1% AI correlation, not emojis proven "more human"
## Example
> **Input:**
> "Excited to share our new feature 🚀 Game-changing AI ✨ Unlock your potential 💡 Let's go 🔥"
>
> **Output (default mode):**
> Score: 78 / 100 — AI-likely
>
> Flagged:
> - 🚀 rocket (3.28% AI freq, 1 use) → consider 📦 or remove
> - ✨ sparkles (3.11% AI freq, 1 use) → remove, no clean swap
> - 💡 lightbulb (2.57% AI freq, 1 use) → 🪛 or remove
> - 🔥 fire (1.31% AI freq, 1 use) → keep or swap to 🌶️
>
> Cluster bonus: +20 (4 distinct AI-pattern emojis)
>
> **Suggested rewrite:**
> "Excited to share our new feature. AI that actually works. Try it. 🌶️"
## Files
- `SKILL.md` — this file
- `../references/emoji-patterns.md` — full frequency table + human-pattern alternatives
## Related skills
- `linkedin-humanizer` — runs after emoji detector to scrub AI vocabulary and rhythm
- `post-audit.md` (sibling) — broader pre-publish check (calls this internally)
- `linkedin-post-writer` — generates drafts that already avoid AI-pattern emojis
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| 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: 472k | Updated | Claude CodeCodex | Verified | Install |
| 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: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |