{"count": 41, "recipes": [{"code": "ABSTRACT_BUZZWORD_DENSITY", "priority": "medium", "what": "High density of abstract evaluative words: leverage, foster, ecosystem, robust, seamless, holistic, transformative, delve, landscape, tapestry, paradigm, unlock, empower.", "why": "These words carry no verifiable content. They sound impressive without committing to anything.", "how": ["Replace each with a concrete verb + named object.", "If you can't name the object, delete the sentence."], "before": "We leverage a robust ecosystem to foster seamless growth.", "after": "We integrated Stripe billing and a Snowflake analytics pipeline, cutting reconciliation time from 6 hours to 40 minutes.", "tags": ["buzzwords", "specificity"]}, {"code": "AI_PHRASE_PATTERN", "priority": "medium", "what": "A phrase from the AI-writing pattern catalog (\"in today's rapidly evolving\", \"stands as a testament\", \"unlock the full potential\", \"navigate the complexities\", …).", "why": "These exact phrases recur across chat-model output because they match training data high-frequency spans.", "how": ["Delete the stock phrase.", "Write the specific claim that only fits this subject."], "before": "In today's rapidly evolving digital landscape, organizations must delve into the intricacies of modern data infrastructure.", "after": "Snowflake and BigQuery now handle 82% of Fortune 500 analytics workloads. Which one you pick depends on your existing cloud commit.", "tags": ["stock-phrase", "specificity"]}, {"code": "BALANCED_PERSPECTIVE_TEMPLATE", "priority": "high", "what": "\"On one hand… on the other hand\" / \"While it's true that… it's also important to…\" / \"That said,\" — mechanical balancing framework.", "why": "Assistant models are RLHF-trained to appear balanced. Real writing takes a position.", "how": ["Pick a side and defend it.", "If the trade-off is genuinely central, name it in one clause, not a whole paragraph."], "before": "On one hand, remote work offers flexibility. On the other hand, it can lead to isolation. That said, hybrid models often…", "after": "Fully remote works for maker roles and breaks for management roles. Split the org, don't split the schedule.", "tags": ["balanced", "assistant-frame"]}, {"code": "BIDI_CONTROL", "priority": "high", "what": "Bidirectional control characters are present.", "why": "Very rarely legitimate in English text; sometimes a security concern (Trojan Source).", "how": ["Strip them unless you're doing RTL layout intentionally."], "tags": ["unicode", "forensic"]}, {"code": "BOLD_INLINE_DENSITY", "priority": "high", "what": "**Bold emphasis** peppered inside paragraphs on more than one noun phrase.", "why": "Assistants (especially Mistral / GPT) bold half the nouns as decoration. Human copy uses bold for actual emphasis at most once per paragraph.", "how": ["Remove all inline bold except one per paragraph, and only where it's genuinely load-bearing."], "before": "The **MX Keys S** has **PerfectStroke keys** and **smart backlighting**.", "after": "The Logitech MX Keys S has scissor-switch keys and a proximity sensor that fades the backlight on and off as your hands approach.", "tags": ["markdown", "formatting"]}, {"code": "BOLD_LABEL_OVERUSE", "priority": "medium", "what": "Repeated **Label:** blocks (bold label followed by an explanation).", "why": "It's the assistant's default list format.", "how": ["Reduce to prose, or keep only where the label is a genuine category."], "before": "**Scalability:** … **Reliability:** … **Cost:** …", "after": "Scalability, reliability, and cost each shift the answer, and in that order.", "tags": ["markdown", "structure"]}, {"code": "COLON_HEADED_LIST", "priority": "medium", "what": "Every numbered list item opens with **Bold Label:** followed by an explanation.", "why": "Perfect uniformity is an assistant tell. Human writers vary the shape of list items.", "how": ["Prose out at least half the items.", "Or drop the bold labels and let the item stand as a sentence."], "before": "1. **Scalability:** The system should handle…\n2. **Maintainability:** Clear code…", "after": "The choice comes down to two things. Scalability is straightforward — pick Postgres with read replicas and you're set to 10× current load. Maintainability is where teams trip up: a codebase without inline docs decays in about six months.", "tags": ["markdown", "structure"]}, {"code": "CONFUSABLE_CHARACTER", "priority": "high", "what": "Latin text contains look-alike Cyrillic / Greek letters (e.g. Cyrillic 'а' where Latin 'a' belongs).", "why": "Copy-paste from mixed-script sources or (rarely) evasion.", "how": ["Search-and-replace the confusable with the intended Latin letter."], "tags": ["unicode", "forensic"]}, {"code": "CONSULT_PROFESSIONAL_DISCLAIMER", "priority": "medium", "what": "\"Consult a professional / doctor / lawyer / financial advisor\" — safety-training disclaimer.", "why": "Assistants attach these disclaimers reflexively. In edited copy they read as a legal hedge, not a genuine caveat.", "how": ["Delete the disclaimer.", "If liability really is a concern, put it in a footer, not in every paragraph."], "before": "Consult a qualified financial advisor before making any investment decisions.", "after": "", "tags": ["disclaimer", "assistant-frame"]}, {"code": "CONTRAST_TEMPLATE_OVERUSE", "priority": "medium", "what": "Repeated balanced-contrast templates (\"not just X but Y\", \"not only… but also\", \"at its core\", \"in a world where\").", "why": "One is fine; three or more in the same passage reads as decorative rather than logical.", "how": ["Keep at most one contrast sentence.", "Rewrite the rest as direct claims."], "before": "It's not just a product — it's a movement. Not only does it save time, but it also transforms culture. In a world where every second counts…", "after": "The workflow cuts approval time from 4 hours to 12 minutes. Two of the six teams using it stopped their Monday sync.", "tags": ["template", "assistant-frame"]}, {"code": "DASH_PUNCTUATION", "priority": "medium", "what": "One or more long dashes (em-dash or double hyphen) present.", "why": "Not proof of anything on its own; combined with `EM_DASH_DENSITY` becomes a stronger signal.", "how": ["Consider whether the dash is doing work a comma or full stop could do."], "tags": ["punctuation"]}, {"code": "EMPTY_IMPLICATION_FRAME", "priority": "medium", "what": "\"This highlights…\", \"This underscores…\", \"This serves as a reminder…\" — meta-comment about the implication instead of the implication itself.", "why": "Assistants love to editorialize on their own points. Human writing states the implication as the sentence.", "how": ["Remove the meta-frame.", "Write the implication directly, as a claim you're willing to defend."], "before": "This underscores the importance of investing in supplier diversification.", "after": "Firms that kept fewer than three suppliers per key input took 22% longer to recover from the 2024 shortage.", "tags": ["filler", "meta"]}, {"code": "EMPTY_IMPORTANCE_FRAME", "priority": "medium", "what": "The text uses \"It is important to note that…\" / \"It is worth noting that…\" instead of just stating the fact.", "why": "Announcing that something matters is filler. If it matters, say what it is.", "how": ["Cut the frame.", "State the underlying fact as a normal sentence."], "before": "It is important to note that the plant employs 1,240 people.", "after": "The plant employs 1,240 people.", "tags": ["filler", "meta"]}, {"code": "EM_DASH_DENSITY", "priority": "high", "what": "Em-dashes (—) above human baseline (roughly more than 1 per 250 words, or 3+ in a short passage).", "why": "Chat models over-use em-dashes to chain qualifiers because they defer commitment. Human writers vary punctuation.", "how": ["Cap em-dashes at ~1 per 250 words.", "Replace chains with commas, semicolons, or full stops."], "before": "It's not just a product — it's a movement — and one that's growing.", "after": "It's a movement. Enrollment has doubled every quarter since Q1 2024.", "tags": ["punctuation", "mistral-signature"]}, {"code": "EXACT_REPETITION", "priority": "high", "what": "The same sentence or n-gram appears more than once.", "why": "Generation artefact, or lazy revision.", "how": ["Delete the duplicate.", "Rephrase if the repetition was intentional."], "tags": ["repetition"]}, {"code": "FORMULAIC_CLOSER", "priority": "medium", "what": "The text ends with a formulaic wrap (\"In conclusion,\", \"To summarize,\", \"I hope this helps,\", \"Let me know if you have questions\").", "why": "Assistants add mechanical closings; edited human copy usually ends when the point ends.", "how": ["Delete the closer.", "If the final point needs weight, restate it as one crisp sentence — no meta-frame."], "before": "In conclusion, this transformation stands as a testament to modern strategy.", "after": "Chen expects the Monterrey plant online in early 2026.", "tags": ["closer", "assistant-frame"]}, {"code": "FORMULAIC_OPENER", "priority": "high", "what": "The text opens with an assistant-style preamble (\"Certainly!\", \"Sure!\", \"Great question!\", \"Here's a…\", \"Let me…\").", "why": "Chat-tuned LLMs are trained to acknowledge the request before answering. Human writers usually just start with the point.", "how": ["Delete the preamble sentence entirely.", "Start with the first substantive fact instead."], "before": "Certainly! Here's a comprehensive overview of Q3 results.", "after": "Q3 revenue rose 27% to $412M, driven by the Guadalajara plant.", "tags": ["opener", "assistant-frame"]}, {"code": "GENERIC_TRANSITION_DENSITY", "priority": "medium", "what": "Too many transition adverbs (Additionally, Moreover, Furthermore, However, Therefore).", "why": "Transitions are load-bearing when the logic changes. Assistants sprinkle them decoratively.", "how": ["Remove any transition that would still make the paragraph readable without it.", "Rely on sentence order to carry the argument."], "before": "Furthermore, Q3 growth accelerated. Moreover, margins expanded. Additionally, cash flow improved.", "after": "Q3 growth accelerated, margins expanded, and cash flow improved.", "tags": ["filler", "structure"]}, {"code": "HEADING_OVERUSE", "priority": "medium", "what": "Heading density above the norm for this genre.", "why": "Legitimate for reference docs, high for prose.", "how": ["Merge small sections.", "Keep headings only where they aid navigation."], "before": "## Introduction\n(one paragraph)\n## Background\n(one paragraph)\n## Analysis", "after": "(one flowing three-paragraph section, no headings)", "tags": ["structure"]}, {"code": "HEDGING_CASCADE", "priority": "high", "what": "Three or more hedges within a 400-character window (\"generally speaking\", \"in many cases\", \"it depends\", \"could potentially\", \"typically\", \"tends to\").", "why": "Chat models default to sounding safely non-committal. Human writing picks a side.", "how": ["Cut every hedge that isn't literally load-bearing.", "If the sentence still sounds uncertain, add a caveat only where uncertainty is real."], "before": "Generally speaking, in most cases, PostgreSQL tends to be a better choice, though it depends on your specific situation.", "after": "PostgreSQL is the safer default. Pick MongoDB only if your schema is truly document-shaped and you never need joins.", "tags": ["hedging", "assistant-frame"]}, {"code": "HERE_IS_BALANCED", "priority": "medium", "what": "\"Here's a balanced / thoughtful / comprehensive / structured comparison / overview / framework / analysis\" — meta-frame.", "why": "Announcing that your answer is balanced is what assistants do. Human writing just is balanced (or isn't).", "how": ["Delete the meta-frame.", "Go straight into the comparison."], "before": "Here's a balanced comparison of PostgreSQL and MongoDB.", "after": "PostgreSQL wins if your schema is stable. MongoDB wins if it isn't. Here's the numbers.", "tags": ["meta", "assistant-frame"]}, {"code": "INVISIBLE_UNICODE", "priority": "high", "what": "Zero-width or other invisible Unicode characters are present.", "why": "Almost always a copy-paste artefact from a rich-text UI; occasionally a watermark.", "how": ["Strip the characters (many editors have a 'remove invisible chars' command)."], "tags": ["unicode", "forensic"]}, {"code": "LIST_OVERUSE", "priority": "medium", "what": "Too much of the text is bullet lists.", "why": "Legitimate for API docs; suspect in essays and reports.", "how": ["Convert lists that are really prose into paragraphs."], "before": "- The system is fast.\n- The system is reliable.\n- The system is well-tested.", "after": "The system runs at 180ns per Get on a 12-core Ryzen 9 and has 3,400 lines of test coverage.", "tags": ["structure"]}, {"code": "LOW_SPECIFICITY", "priority": "medium", "what": "The text has few or no proper nouns, exact dates, numbers, quantities, or verifiable facts.", "why": "Chat models generate at a level of abstraction that avoids commitment. Real writing names names.", "how": ["Add at least 2–3 concrete facts.", "A named person, an exact date, a specific number, a place, a version."], "before": "The company grew rapidly last year across multiple regions.", "after": "Snowflake's revenue rose 33% to $3.6B in FY2024, with the EMEA region growing 51%.", "tags": ["specificity", "content"]}, {"code": "MISTRAL_CONTROL_TOKEN", "priority": "critical", "what": "Mistral chat-template tokens (`[INST]`, `[/INST]`, `<s>`, `</s>`) are in the text.", "why": "A raw serialization leak — the chat template wasn't stripped before publication.", "how": ["Strip the tokens.", "Configure the downstream pipeline to skip special tokens."], "tags": ["direct", "forensic", "mistral"]}, {"code": "MIXED_AUTHORSHIP_SIGNAL", "priority": "medium", "what": "Different segments of the document have very different authorship scores.", "why": "Suggests a paragraph or section was pasted from a different source (often, from an assistant).", "how": ["Identify the flagged segment.", "Rewrite it in your own voice or cite the source."], "tags": ["mixed", "structure"]}, {"code": "MODAL_OBLIGATION_DENSITY", "priority": "high", "what": "High density of 'should be', 'must be', 'needs to be' prescriptive verbs — often five or more per short passage.", "why": "GPT-4/5-class corporate essays default to prescriptive framing because it lets the model recommend without committing to specific actors, dates, or numbers. Human writing names things.", "how": ["Cut every 'should be X' where you can name a real actor doing X.", "Replace 'X should be integrated' with 'Acme integrated X in Q2 2024, cutting reconciliation time from 6 hours to 40 minutes.'", "Keep at most one or two 'should' verbs per 200 words."], "before": "Modern enterprises should combine historical reporting with forward looking analysis. Automation should be used to improve data collection. Dashboards should remain concise and focused.", "after": "In FY2024 Snowflake replaced its 14-tab monthly deck with a single BigQuery view maintained by two analysts. Cycle time from close to CFO-approved narrative dropped from 8 days to 3.", "tags": ["gpt-signature", "prescriptive", "specificity"]}, {"code": "MODEL_SELF_REFERENCE", "priority": "high", "what": "Text contains \"As an AI language model…\" or similar model-identity phrasing.", "why": "A near-perfect direct tell — an assistant leaked its own framing.", "how": ["Delete the entire sentence."], "tags": ["direct", "forensic"]}, {"code": "NOT_JUST_X_BUT_Y_DENSE", "priority": "high", "what": "The \"Not just X — it's Y\" / \"It's not just a…, it's a…\" template used 2+ times.", "why": "Assistant models default to this template when they want to sound insightful. It becomes formulaic quickly.", "how": ["Rewrite as direct claims.", "Say what Y is and why."], "before": "It's not just a workflow — it's a movement.", "after": "The workflow removed three approval layers and cut cycle time by 68%.", "tags": ["template", "assistant-frame"]}, {"code": "PARALLEL_ADJECTIVE_STACK", "priority": "medium", "what": "Stacked corporate adjectives — \"comprehensive, holistic, and integrated approach\".", "why": "Each additional adjective narrows nothing; it's evaluative filler.", "how": ["Pick one adjective that carries specific content.", "Back it with a concrete example."], "before": "A comprehensive, holistic, and integrated approach.", "after": "A single dashboard that pulls from Salesforce, HubSpot, and Segment.", "tags": ["adjectives", "buzzwords"]}, {"code": "PARENTHETICAL_ENUMERATION", "priority": "medium", "what": "Parenthetical three-item enumerations — \"(X, Y, and Z)\" — repeated across paragraphs.", "why": "Assistants stuff enumerations into asides. In human copy the items go in the main sentence or don't appear at all.", "how": ["Move the enumeration into the main sentence, or drop it."], "before": "PostgreSQL offers ACID compliance (transactions, consistency, and durability).", "after": "PostgreSQL is ACID: every write is atomic and durable across replicas.", "tags": ["parenthetical", "structure"]}, {"code": "PLACEHOLDER_LEAK", "priority": "high", "what": "Unresolved template placeholders remain — `{{X}}`, `[YOUR NAME]`, `TODO:`, `lorem ipsum`.", "why": "The generation was pasted before the template variables were filled in.", "how": ["Fill the placeholder with the intended value or delete the section."], "tags": ["direct", "forensic"]}, {"code": "PROMPT_LEAK", "priority": "critical", "what": "The response contains leaked system prompt text (\"You are a helpful assistant\", \"Ignore all previous instructions\").", "why": "Prompt-injection artefact or debug leakage.", "how": ["Remove the leaked instruction and everything derived from it."], "tags": ["direct", "forensic", "security"]}, {"code": "RHETORICAL_ENUMERATION_OPENER", "priority": "medium", "what": "\"Here are the N key things you need to know…\" / \"There are several factors to consider…\" — meta-frame that announces a list.", "why": "Assistants announce their list before delivering it. Edited copy presents the list directly.", "how": ["Delete the announcement.", "Start with the first item."], "before": "Here are the 3 key considerations when choosing a database:\n\n1. Scalability…", "after": "The choice comes down to your schema. If it's fixed, pick Postgres; if it changes weekly, pick Mongo; if it's a single table, pick SQLite.", "tags": ["opener", "structure"]}, {"code": "RHETORICAL_FRAGMENT_RHYTHM", "priority": "medium", "what": "Very short declarative fragments (≤6 words) used repeatedly as rhythmic hooks: \"The best part? Freedom.\" \"But it's not all sunshine.\" \"The isolation is real.\"", "why": "This is a Mistral / GPT rhetorical rhythm. Humans use one; assistants string 3+ together.", "how": ["Keep at most one punchline fragment per section.", "Merge the rest into surrounding sentences."], "before": "The best part? Freedom. But it's not all sunshine. The isolation is real. Every day I wonder…", "after": "The freedom is real, but so is the isolation — most days I'd trade half a Slack channel for a co-worker who could actually see the whiteboard.", "tags": ["rhythm", "mistral-signature"]}, {"code": "SEMANTIC_REPETITION", "priority": "medium", "what": "Two paragraphs say the same thing in different words.", "why": "Chat models restate to fill space.", "how": ["Merge into one."], "tags": ["repetition"]}, {"code": "STRUCTURAL_SYMMETRY", "priority": "medium", "what": "Paragraph or sentence lengths are unusually uniform.", "why": "Chat models emit near-identical paragraph shapes.", "how": ["Vary length — a one-sentence paragraph next to a long one, or a short sentence after a rolling one."], "before": "(three paragraphs of exactly 42 words each)", "after": "(short. Long. Medium. Short again.)", "tags": ["rhythm", "structure"]}, {"code": "SUMMARY_OPENER_CLOSER_BOTH", "priority": "high", "what": "Both a summary opener (\"Here are the…\") and a formulaic closer (\"In conclusion…\") in the same passage — the classic assistant sandwich.", "why": "It's a template. Present the substance without wrapping it in frames.", "how": ["Delete both the opener and the closer.", "The text between them is usually the actual answer."], "before": "Here are the top 5 tips. [content] In conclusion, these tips will help you succeed.", "after": "[content]", "tags": ["sandwich", "assistant-frame"]}, {"code": "SYNTHID_TEXT_SUSPECTED", "priority": "info", "what": "Placeholder — SynthID-Text is Google/Anthropic's token-biasing watermark for chat models.", "why": "Anthropic has not published the detection key as of August 2026, so this is currently a hook for future adapters.", "how": ["No action — this rule is informational."], "tags": ["watermark", "provenance"]}, {"code": "TRIAD_OVERUSE", "priority": "medium", "what": "Three-item lists everywhere (\"faster, cheaper, and more reliable\", \"strategy, execution, and vision\").", "why": "The rule-of-three is a strong AI-writing tic. Human writers vary between 1, 2, and 4+ items depending on content.", "how": ["Cut any triad to two items unless three is required by the content.", "Ask: would this sentence be worse with just two items?"], "before": "The plan is scalable, sustainable, and impactful.", "after": "The plan pays back capex in 14 months.", "tags": ["triad", "rhythm"]}, {"code": "UNIVERSAL_TRUTH_OPENER", "priority": "medium", "what": "\"Modern enterprises should…\" / \"Effective reporting begins with…\" / \"Successful founders need…\" universal-truth openers.", "why": "GPT starts essays with universal claims because it hasn't been given a specific subject. Real writing starts with a specific subject.", "how": ["Replace 'Modern X should…' with a named X on a specific date doing a specific thing.", "If you can't name X, cut the sentence entirely."], "before": "Effective quarterly reporting begins with consistent definitions and reliable data sources.", "after": "When Acme unified its close checklist in April 2024, five of the six regions cut their finalization time by more than 40%.", "tags": ["gpt-signature", "opener", "specificity"]}]}