Slop*Detector

CERTIFIED SLOP — slop score 79/100

KrillCLaw's landing page hits nearly every mark on the vibe-coded checklist, and the 79/100 score reflects real density rather than a few stray tells. The green-on-black terminal theme with its neon glow halo around the mascot is the kind of aesthetic default a model reaches for when asked to make something look techy, and it's paired with a 3x3 feature grid ('Secure by Default,' 'Every Channel,' 'Real Memory') that reads as slots filled from a template rather than a considered layout. The copy compounds it: em dashes show up in the illustrative examples, punch fragments like 'No app. No screen.' sit right next to a full drumbeat of clipped sentences, and the closing lines lean on the banned 'X, not Y' reframe twice. Add invented competitor stats with bare question marks next to KrillCLaw's own suspiciously rounded numbers, plus a hero stat ('<59KB') immediately undercut by '407KB' in the fine print, and you get a page that never seems to have had a single editor read it end to end. This one is about as clear-cut a certification as the tool produces.

Findings

• Green-on-black terminal theme — The hacker-terminal look (black background, green accents, code-style monospace body text) is a top-three vibe-coded aesthetic, especially for dev tools.
• Emoji-prefixed headings — Feature headings that open with 🚀 or ✨ are generated-copy garnish.
• Green-on-black terminal theme — The entire page uses a near-black background with phosphor-green accent text and neon glow around the logo, plus monospace-feeling headers. Per the guide this specific combination is a low-confidence, minor signal on its own but adds to the density of tells here.
• Neon/glow effects — Unmotivated glowing halo effects appear around the hero image border, which the guide explicitly calls out as a confirmed AI tell ('AI loves this glowing shit for no reason').
• Feature-tile grid — A 3x3 grid of identical cards ('Secure by Default', 'Every Channel', 'Real Memory', 'Tested & Leak-Free', 'Skills & Plugins', '20+ LLM Providers', 'MCP Compatible', '7 Channels') each with bold subheader and two-line blurb — the canonical generated feature-grid block.
• Template skeleton / stock section sequence — The page follows the familiar stock landing-page section order: hero with big stat, feature explainer cards, feature grid ('Built Different'), architecture/comparison tables, 'How it compares' competitor table, code sample — a template-shaped assembly of generic SaaS sections.
• Redundant sections — Multiple sections restate the same Lite/Full distinction in different formats (cards, big numbers, a table, a use-case grid), which reads as template slots being filled rather than a single well-organized explanation.
• Pill-style badges — Small fully-rounded badge chips are scattered across the use-case grid and comparison sections labeling items as LITE, FULL, or BOTH — matching the pill-tag tell (though below the 10+ threshold for maximum severity).
• Hero/body coherence check (stat inflation) — The hero centers dramatically on a huge '<59KB' stat, but this is immediately undercut by 'full: 407KB' directly below it — the headline stat oversells a number that doesn't represent the actual full product, a classic case of a generated hero not knowing what the body content says.
• Invented/overprecise competitive stats — The 'How it compares' table invents suspiciously exact competitor stats (some marked with bare '?' for unknown fields) alongside self-favorable rounded claims, a pattern consistent with fabricated competitive benchmarking common in generated marketing copy.
• Em dashes — Em dashes are the single most reliable punctuation tell of LLM copy. Human marketing copy almost never uses them; models reach for them constantly.
• One/two-word punch sentences — Single dramatic words dropped as their own sentence ("Done." "Gone." "Period.") are an LLM emphasis habit, not a human one.
• Stacked short-sentence drumbeat — Three or more clipped sentences in a row ("Important emails buried. Drafts that sound wrong. Labels everywhere.") is the manufactured-rhythm tell.
• Sparkle/rocket emoji — ✨ and 🚀 scattered through headings and feature lists are the default garnish of generated landing copy.
• Em dashes present — The guide bans em dashes entirely; several appear in dramatic mid-sentence pauses inside the illustrative examples, a classic LLM punctuation tic.
• Appositive 'X, not Y' reframe — Two separate closing lines use the banned reframe-in-miniature construction to deliver a punchy contrast instead of stating the point directly.
• Duplicated template sections — Two different sections ('What becomes possible') use near-identical framing and an identical closing formula, suggesting the copy was generated from a repeated template rather than written fresh for each section.
• Short fragment stacking — A stack of two-to-three-word declarative fragments builds rhythm rather than flowing prose, matching the banned drumbeat pattern.
• 'Real' used as intensifier — The word 'real' is used to manufacture sincerity/significance rather than being earned by specifics, which the guide flags as a crutch.
• Rule-of-three construction — A triad of parallel 'no' clauses is used for rhythmic emphasis, a mild instance of the compulsive rule-of-three the guide warns against.