CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

nl-humanizer-ai-smell-detector

Detect and remove "AI smell" from Dutch text with a diagnostic scoring system and human-centric rewriting. Use when: Audit Dutch marketing copy for robotic linguistic patterns; Convert formal Netherlands-Dutch into conversational Belgian-Dutch; Reduce AI detection scores through targeted linguistic rewriting.

Quality

69%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./plugins/nex-business-operations/skills/nl-humanizer-ai-smell-detector/SKILL.md

The canonical home for this skill is nl-humanizer-ai-smell-detector in NexaiGuy/nex-ai-skills

SKILL.md
Quality
Evals
Security

Quality

Content

42%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body reads as a marketing landing page rather than an operational skill: it convincingly describes a two-phase detect-and-rewrite system but omits the actual detection rules, scoring thresholds, and rewriting instructions it promises, since the original scripts and templates are absent. Nearly half the content (FAQ, "Why use this skill", promo links) is promotional padding that restates the description.

Suggestions

Replace the descriptive "What it does" prose with the operational core: the concrete detection rule list (flagged words/patterns and how they weight the 0-5 AI Smell Score), the score thresholds, and the rewriting rules — either inline or in a bundled references file linked from SKILL.md.

Delete the "Questions" FAQ section and the "Why use this skill" marketing bullets; they restate the frontmatter description and add promotional padding (~30 lines that earn no instructional value).

Add an explicit validation loop to the workflow: after rewriting, re-run the linguistic audit and report the before/after AI Smell Score to confirm the transformation worked while preserving the original intent.

DimensionReasoningScore

Conciseness

The ~25-line "Questions" FAQ restates "What it does" and "When to use" with marketing padding ("designed to integrate seamlessly with your existing AI agent workflows", "ensuring your content resonates with local audiences"), and "Why use this skill" bullets plus footer promo links are promotional rather than instructional. Not 1 because it does not explain concepts Claude already knows; not 3 because the redundancy and padding span multiple sections rather than being minor tightenings.

2 / 5

Actionability

Concrete tells are named ("*uiteraard*, *desalniettemin*, *tevens*", "het is belangrijk om", passive constructions, tricolon structures) and a worked input-to-score-to-rewrite example is provided, but the promised "deterministic diagnostic tool" scoring logic and rewriting rule set are absent ("Scripts, extension source, and templates from the original package are not in this folder"). Not 2 because the tells list and worked example are usable concrete guidance; not 4 because no executable rule set or procedure exists.

3 / 5

Workflow Clarity

The two-phase process is stated descriptively ("The skill operates in two phases: detection and transformation") but not as an operational sequence — there are no scoring criteria, no explicit checkpoints, and no validation loop such as re-scoring after rewriting to confirm the AI smell was reduced. Matches anchor 3: sequence present but checkpoints missing or implicit.

3 / 5

Progressive Disclosure

Sections are clearly headed with no nested references, but roughly 30 lines of FAQ and marketing content are inlined that add no instructional value, and the detailed materials the skill claims (scripts, templates, rule lists) have no home — the body itself admits "Scripts, extension source, and templates from the original package are not in this folder." Not 2 (short and organized, not a monolithic wall); not 4 (non-instructional content is inlined and the promised depth simply does not exist anywhere).

3 / 5

Total

11

/

20

Passed

Description

80%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A well-formed description with an explicit 'Use when:' trigger list, a distinct Dutch-language niche, and several concrete capabilities. Its main weakness is trigger-term coverage: the most natural user phrasings ("humanize", "make it sound natural", "Nederlands", "Vlaams") are missing in favor of branded jargon like "AI smell".

DimensionReasoningScore

Specificity

"Detect and remove 'AI smell' from Dutch text with a diagnostic scoring system and human-centric rewriting" plus "Audit Dutch marketing copy" and "Convert formal Netherlands-Dutch into conversational Belgian-Dutch" list several concrete actions. Falls short of 5 because "human-centric rewriting" and "Reduce AI detection scores" stay generic about the mechanism, leaving minor coverage gaps; above 3 because more than 1-2 specific actions are named.

4 / 5

Completeness

Explicitly answers both: a clear "what" ("Detect and remove 'AI smell' from Dutch text with a diagnostic scoring system and human-centric rewriting") and an explicit "Use when:" clause with three concrete trigger phrases. Matches the anchor-5 example pattern; anchor 4 requires a 'when' that could be more explicit, which is not the case here.

5 / 5

Trigger Term Quality

Relevant keywords are present ("Dutch marketing copy", "Belgian-Dutch", "AI detection scores") but the natural phrases users would most likely say — "humanize", "make text sound natural", "Nederlands", "Vlaams" — are missing, and "AI smell" is branded jargon a user would not spontaneously use. Not 2 (more than a couple of generic keywords); not 4 (central natural phrasings and synonyms are absent).

3 / 5

Distinctiveness Conflict Risk

A clear niche (Dutch-language AI-smell detection and Belgian-Dutch localization) with distinct, language-specific triggers ("Netherlands-Dutch", "Belgian-Dutch", "Dutch marketing copy") that virtually no other skill would match. Minimal conflict risk.

5 / 5

Total

17

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

NexaiGuy/nex-ai-skills

Path

plugins/nex-business-operations/skills/nl-humanizer-ai-smell-detector/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.