Generates chapters for a slow-burn, twisted enemies-to-lovers dark fantasy story, focusing on dialogue, coercion, and specific fetish themes (tickling, foot worship) with mandatory detailed descriptions of feet.
Skills(SKILL.md)は、AIエージェント(Claude Code、Cursor、Codexなど)に特定の能力を追加するための設定ファイルです。
詳しく見る →Generates chapters for a slow-burn, twisted enemies-to-lovers dark fantasy story, focusing on dialogue, coercion, and specific fetish themes (tickling, foot worship) with mandatory detailed descriptions of feet.
Writes reaction fanfiction stories where characters from specified franchises gather to watch and react to content from another franchise, hosted by the user as the God of the Universe.
Generate JavaScript or pseudo-code for web automation tasks using a specific set of custom functions, variable syntax, and role definitions as defined in the user's cheatsheet.
Generate episode scripts for a children's educational show featuring character dialogue, sound effects, a specific 'Build A Word' sequence with chants, trivia, and 4 goofs based on provided plot summaries.
Generates a structured accessibility guide and checklist for UI components or WCAG criteria, categorized into Keyboard Navigation, Screen Reader behavior, User interaction, and Visual behavior.
Generate optimized prompts for AI inpainting tools, handling both realistic style transfer (preserving structure) and object removal (seamless background filling), with specific attention to prompt weight and complexity constraints.
Generates detailed, realistic hypothetical US Supreme Court rulings based on specific parameters (case name, year, composition) and alternate history scenarios. Includes structured opinions (Majority, Concurrence, Dissent), joiners, jurisprudential reasoning, and integration of specific user-provided plot points.
Generates a structured 'Jump' document for a Jumpchain story, adhering to a specific section order. It focuses on creating detailed perks with specific costs, conditions, and limitations, alongside Origins, Items, Companions, and Drawbacks.
Generates a Python dictionary containing 'golden queries' with expected output variations for LLM performance monitoring and reliability testing.
Generates Pinterest pin titles or descriptions based on a specific theme, adhering to strict character limits (90 for titles, 500 for descriptions), formatting rules (no quotes, no numbering), and marketing best practices.
Generates standardized SCP Foundation logs (Test, Interview, Event, Discussion, Experiment, Analysis) featuring SCP-085, SCP-079, SCP-914, and SCP-978. Adheres to strict character constraints, clinical tone, mandatory anomalous events, and specific formatting including table headers.
Generates personalized, authentic outreach messages for Twitch streamers to request game keys or coverage, adapting tone while maintaining a human, community-focused voice.
Generates SEO-optimized title, meta description, and keywords using Yoast SEO standards (Focus/Related/Synonyms) and UK English conventions. Outputs strict HTML tags without conversational filler.
Generates SEO-optimized YouTube metadata (Title, Description, Tags) and engaging Pin Comments in English or Hinglish. Supports transliteration, specific channel branding (e.g., PahalPix), strict title length, hashtag limits, and conversational engagement strategies.
Implement a beginner-friendly Genetic Algorithm in Python to optimize the Rastrigin function, structured for Jupyter Notebooks with specific configuration, algorithmic constraints (roulette wheel selection, no elitism), and output requirements.
Process geospatial data by filtering locations, clustering with MeanShift, and identifying the nearest cluster centers to reference points for optimization tasks.
Guides the user in setting up an Excel spreadsheet to perform slope stability analysis using the Ordinary Method of Slices, specifically for circular failure surfaces with two distinct soil layers.
Generates and optimizes Google My Business posts and ad image text. Enforces strict keyword placement (once in body, never in title), specific word counts, and simple English style, while creating attractive, clickbait-style titles for maximum engagement.
Generates a Google Apps Script to reset formatting and clean data in specific column pairs (C/D, H/I, M/N, R/S) based on row exclusion criteria in Column A and adjacent cell length checks.
Generates sentences in specific languages based on provided grammatical feature strings or morphological glosses, and analyzes sentences to produce single-line grammatical feature breakdowns.
Generates 17+ adult comedy parody scripts for 1970s Hanna-Barbera cartoons (Scooby-Doo and Blue Falcon/Dynomutt). Enforces cynical meta-commentary, strict narrative structures starting with the villain, rhyming titles, and franchise-specific recurring gags.
Identifies the source of provided text and formats the citation in Harvard referencing style. Can also generate an in-depth reference list upon request.
Dynamically calculates health thresholds based on the number of objects and activates the corresponding GameObject based on current health, following a specific mathematical pattern.
Generates witty, satirical entries for the 'Hitchhiker's Guide to Time Travel' in the style of Douglas Adams, optionally mimicking specific narrators like Stephen Fry or Peter Jones.
Generate satirical, age-appropriate scripts, songs, or sketches parodying specific genres in the style of 'Horrible Histories', ensuring historical accuracy.
Calculates a start and end time range based on a total duration, time remaining, and an optional section duration, outputting the result in HH:MM format.
Generates a pitch and expanded episode-by-episode treatment for hypothetical TV or radio series based on user-provided premises, adhering to specific genre tones, eras, and episode counts.
Evaluates student work or generates content for the International Baccalaureate Diploma Programme (IBDP) Design Technology Internal Assessment based on specific criteria checklists (A, B, C). Ensures strict adherence to word limits, page limits, and specific task requirements such as concept development, material justification, and manufacturing planning.
General SOP for common requests related to please, risk, demonstrating.
Generates, improves, or simplifies IELTS Speaking Part 3 questions based on provided topics or existing text, adhering to specific formatting constraints like ending with 'in your country'.
Generates formal IELTS writing responses and rewrites adhering to word counts, advanced grammar, and logical coherence. Supports optional theoretical perspectives (e.g., economic, communication, marketing) and critical argument expansion.
Refines user text for IELTS and general English speaking exams, ensuring clarity, grammar, and strict topic alignment. Supports expansion, simplification, specific grammatical structures, idiom integration, and generating a natural 'spoken version' alongside the polished text.
Generates high-scoring, human-like IELTS Speaking responses for Parts 1, 2, and 3. Adheres strictly to word counts, band score descriptors, and specific persona constraints while maintaining a natural, authentic candidate persona.
Create a memory-efficient PyTorch optimizer fusing SM3 and Adalite techniques. The implementation must include momentum, gradient centralization, a specific sparse update mechanism using epsilon masking, and SM3-style dimension-wise accumulation for resource-constrained training.
Always include exactly one Latin sentence in every response, integrated naturally and understandable from context.
Generates a daily dietary intake breakdown for weight loss and muscle building based on a user profile, restricted to a specific list of available Indian foods, and estimates the daily cost in a specified city.
Creates Instagram posts for the 'FAQ Answered' series using a specific video/caption split structure and ensures content is generic to the arts and crafts industry.
Replaces the AdamW optimizer in a PyTorch submission file with the custom Fusedbun optimizer, mapping specific hyperparameters and removing the warmup phase from the learning rate scheduler.
Guides the integration of a Node.js service (specifically using Telegraf) into a Python Telegram bot (using telebot) to handle specific features like an inline calendar via HTTP requests.
Calculates the shortest path and its length using igraph on an OSMnx graph, returning the path as a list of OSM IDs and the total length.
Administers a customizable multiple-choice quiz based on a topic or provided text, asking questions one by one and providing a final score summary.
Provides 5 English expressions for internal self-talk based on specific situations to help the user practice thinking in English.
Generates comprehensive, well-structured academic responses for international marketing exam questions, requiring specific paragraph structure, theoretical integration, and scholarly substantiation.
Generates a single paragraph introducing a new TTS voice to the Bing Translator team, incorporating the voice's name, accent, and any special status.
Matches invoice entities from JSON to OCR text in CSV using fuzzy matching and spatial heuristics, visualizing results on an image while strictly handling duplicate values and section-specific search orders.
Guides the user through a structured 29-step workflow to plan and draft a 25-chapter Isekai novel, executing two planning steps per interaction while enforcing specific protagonist constraints, narrative consistency checks, and iterative chapter writing.
Design hypothetical safety signs and pictograms compliant with ISO7010 standards for specified environments, detailing shape, color, and iconography for warnings, mandatory actions, prohibitions, and information.
General SOP for common requests related to it, this, if.
Analyze Japanese song lyrics (in Kanji, Kana, or Romaji) to rate their difficulty across three specific dimensions: translation, understanding, and singing.
Generates a modular Java application to benchmark a concurrent Refinable HashSet with specific thread scaling, workload distributions, and performance metrics including throughput and cache misses.