jwynia/agent-skills
118 skills56.1k installs15.9k starsGitHub
Install
npx skills add https://github.com/jwynia/agent-skillsSkills in this repo
1Frontend DesignThe frontend-design skill combines design philosophy with Deno scripts to analyze existing CSS and generate palettes, typography, tokens, and framework components that avoid generic AI aesthetics. Core principle demands a clear aesthetic direction executed with precision, whether maximalist or minimal. Three modes cover Analyze for style and accessibility audits, Specify via JSON specs in assets, and Generate for palettes, typography scales, tokens, and React Vue Svelte or HTML components. Prerequisites require Deno with allow-read and allow-write permissions. Analyze scripts inventory colors, typography, spacing inconsistencies, extract tokens to css scss tailwind or style-dictionary formats, and check WCAG AA or AAA contrast, focus, touch targets, and motion preferences. Generate scripts support seed colors, themes warm cool neutral vibrant muted dark light, type scales from minor-second through golden-ratio, and component templates with aesthetic variants minimal bold organic brutalist glassmorphism or neumorphism. Skip for backend-only work or simple text edits. Teams invoke it when building web UIs, auditing legacy CSS, or bootstrapping design systems with intentional visual.2.3kinstalls2Requirements AnalysisRequirements Analysis diagnoses requirements-level problems and guides discovery from vague intent to validated needs without premature solution thinking. Six states progress from RA0 no problem statement through solution-first thinking, vague needs, hidden constraints, scope creep prevention, to RA5 validated requirements ready for system design handoff. Each state lists symptoms, key questions, and interventions such as Jobs-to-be-Done interviews, function extraction without technology words, specificity ladders, constraint inventories, and MoSCoW prioritization. Core principle treats requirements as hypotheses about what solves a problem rather than documents to freeze. Diagnostic process listens for symptoms, starts at the earliest problem state, applies interventions, validates indicators, and produces artifacts from templates. Anti-patterns flagged include solution specifications, stakeholder fiction without evidence, infinite backlogs, and premature precision on unvalidated assumptions.2.2kinstalls3Godot Best PracticesGuide AI agents in writing high quality GDScript code for Godot 4 x This skill provides coding standards architecture patterns and templates for game development Use this skill when Generating new GDScript code Creating or organizing Godot scenes Designing game architecture and node hierarchies Implementing state machines object pools or save systems Answering questions about GDScript patterns or Godot conventions Reviewing GDScript code for quality issues Do NOT use this skill when Working with C in Godot use C patterns Working with Godot 3 x syntax differs significantly Using GDExtension C different paradigm Working with Godot s visual scripting Follow GDScript naming standards consistently gdscript Classes PascalCase class_name PlayerController extends CharacterBody2D The godot best practices agent skill provides documented workflows prerequisites triggers and safety guidance from its SKILL md source Agents load it when user requests match the description and follow step by step instructions without inventing capabilities It integrates with standard agent tooling for the tasks inputs outputs and failure modes described in the repository documentation1.9kinstalls4Presentation DesignThe presentation-design skill provides diagnostic and assistive frameworks for presentations that communicate effectively across reveal.js, PowerPoint, Keynote, and Google Slides. Core principle is audience-centered design where every decision serves audience understanding not presenter convenience. Phase one covers audience persona, one-sentence main message, three to five supporting points, evidence mapping, call to action, and essential versus expandable content under time constraints. Phase two replaces bullet walls with assertion-evidence structure pairing complete sentence assertions with supporting visuals. Quick reference table maps wall of text, bullet disease, kitchen sink, pretty but empty, and cognitive overload symptoms to fixes. Later phases address visual strategy, slide layout, delivery preparation, and evaluation rubrics. Applicable to any presentation tool with message-first design emphasis. Use when designing presentations, creating slides, structuring talks, getting presentation feedback, or reviewing slide decks for clarity.1.8kinstalls5Web Search TavilyThe web-search-tavily skill search the web using Tavily API for high-quality AI-optimized results with advanced filtering options Use when you need structured search results domain filtering relevance scores or AI-generated answer summaries Requires TAVILY_API_KEY Keywords tavily advanced search filtered search domain filtering relevance scoring Web Search Tavily API Search the web using Tavily's AI-optimized search API Returns high-quality structured results with relevance scores and optional AI-generated summaries Note This skill requires a Tavily API key For basic web search using the agent's built-in capability see web-search When to Use This Skill Use this skill when You need to find current information not in your training data The user asks about recent events news or updates You need to verify facts or find authoritative sources Research requires real-time web data Keywords mentioned search look up find online current latest news Do NOT use this skill when Information is already in your knowledge base and doesn't need verification The user asks about historical facts that don't change You're working with local files1.1kinstalls6Godot Asset Generatorgodot-asset-generator is a Deno-based agent skill in jwynia/agent-skills that turns AI image generation into Godot-ready game art. It covers concept art, style guides, sprites, sprite sheets, tilesets, UI icons, and animation frames using providers such as DALL-E, Replicate, and fal.ai, then writes import configuration for Godot 4.x. Developers reach for godot-asset-generator when they need repeatable 2D asset batches without hand-painting every frame or manually tuning Godot import presets. API keys such as OPENAI_API_KEY, REPLICATE_API_TOKEN, or fal.ai credentials are required for the chosen provider.909installs7Pdf Generatorpdf-generator is an MIT-licensed agent skill (version 1.0) that creates and manipulates PDF files programmatically from agent workflows. It supports template-based generation including form filling and overlays, plus from-scratch PDF creation, merging multiple documents, adding watermarks, and extracting text content. The skill runs on Deno and requires --allow-read and --allow-write permissions for filesystem access. Developers reach for pdf-generator when an agent must output invoices, filled government forms, merged report packets, or watermarked drafts without leaving the coding session. Keywords in the manifest include PDF, form, fillable, merge, watermark, extract, and report.901installs8Story CoachStory Coach is an assistive writing agent skill that acts as a strict non-generative coach for fiction, scripts, and narrative content. Instead of producing prose, dialogue, scenes, or outlines, it asks probing questions, diagnoses problems in the writer's existing draft, explains relevant storytelling frameworks, and suggests approaches the writer can execute themselves. The skill is built around the core belief that the writer knows their story best and that guided discovery produces more authentic, lasting results than having an AI write on their behalf. It is ideal for indie creators, novelists, screenwriters, and game narrative designers who want high-quality coaching while protecting their own creative ownership and voice.812installs9Document To Narrationdocument-to-narration is a Claude Code skill from jwynia/agent-skills that transforms long-form written material—documentation, reports, or book chapters—into narration scripts suitable for podcasts, audiobooks, or agent voice synthesis. Developers reach for document-to-narration when they have dense technical or narrative prose that must become spoken-word content without hand-editing every paragraph for audio pacing and clarity. The skill automates the structural rewrite from reading format to listenable format, preserving meaning while adapting sentence rhythm, transitions, and emphasis for audio delivery. It fits documentation pipelines, content repurposing workflows, and agent systems that need text-to-speech or voice-output scripts generated from existing markdown or report files.765installs10Web Searchweb-search is a MIT-licensed agent skill at version 1.0 from jwynia/agent-skills that instructs compatible agents to query the web through the built-in WebSearch capability. No external API keys are required, and the skill documents compatibility with any agent that exposes WebSearch. Developers invoke web-search when training data may be stale—current events, release notes, pricing changes, library deprecations, or factual verification during research sessions. The skill metadata classifies it as a generative utility in the research domain and lists trigger keywords including search, lookup, find, research, current events, and facts. It is a lightweight orchestration pattern for standardized web retrieval rather than a standalone search service.756installs11Electron Best Practiceselectron-best-practices is a version 1.0 agent skill guiding secure Electron 20+ desktop app development with React, electron-vite, and Electron Forge packaging. The skill enforces context isolation, sandbox mode, and disabled nodeIntegration, routing all main-renderer communication through contextBridge invoke/handle patterns with typed IpcChannelMap definitions. Bundled asset templates cover electron.vite.config.ts, forge.config.js, Playwright config, typed IPC handlers, preload scripts, and multi-window examples. Developers reach for electron-best-practices when scaffolding production desktop apps requiring code signing, notarization, CSP headers, and Playwright end-to-end tests. The skill explicitly excludes Tauri projects and Electron versions below 20 where security defaults differ.734installs12Docx Generatordocx-generator is an agent skill from jwynia/agent-skills backed by a Deno analyze-template.ts script that reads DOCX files with JSZip (npm:jszip@3.10.1) and JSR std libraries. It extracts paragraphs, tables, headers, footers, and placeholders from Word templates, emitting JSON structure for replacement planning before generating or populating new documents. The workflow uses deno run --allow-read scripts/analyze-template.ts with options for verbose and pretty JSON output. Developers reach for it when building report generators, mail-merge style tools, or agent pipelines that must inspect .docx layout programmatically rather than manually editing Word. It suits Deno-based automation where filesystem read access to template files is available.719installs13Summarizationsummarization is a version 1.0 MIT utility skill by jwynia that treats summarization as translation rather than mere word reduction. It matches summary type to purpose, audience, and context when users ask to summarize, create a TLDR, condense content, or produce executive summaries and abstracts. Instead of a one-size-fits-all shorten pass, the skill selects the right structure—quick TLDR, decision-oriented executive brief, or reference digest—based on how the output will be used. Developers reach for summarization when ingesting long RFCs, research papers, incident postmortems, or product specs that need a stakeholder-ready distillation without losing decision-critical nuance.669installs14WorldbuildingWorldbuilding is an agent skill that diagnoses world-level story problems in fiction. It analyzes settings that feel like painted backdrops, technology or magic that never transformed society, institutions that feel arbitrarily designed, economies without internal logic, and non-human species that act like humans in costume. The skill identifies which of five progressive world states your story currently occupies and supplies concrete, specific interventions that create the perception of independent history, systemic consequences, and emergent depth. It is especially useful for novelists, game writers, and world-driven content creators who sense something is missing but cannot pinpoint the exact worldbuilding fractures. Because it follows the core principle that worlds fail when they feel designed rather than evolved, it steers creators away from cosmetic additions toward systemic coherence that operates independently of the plot.639installs15System Designsystem-design is a version 1.0 MIT-licensed agent skill from jwynia/agent-skills that helps developers translate validated requirements into architecture decisions, component designs, and interface definitions. The skill treats design as constraint-driven trade-offs, diagnosing integration gaps and over-engineering risks before code changes accumulate. Developers reach for system-design when requirements are approved but service boundaries, data flows, and component responsibilities are still ambiguous. The assistive diagnostic mode guides explicit decision records rather than prescribing a single blueprint, making outputs useful for backend services, APIs, and multi-component applications where missed integration points are costly.615installs16Novel RevisionNovel Revision is an agent skill that helps writers manage revisions across multiple levels of abstraction in long-form fiction. It operates on three simultaneous layers: thematic/conceptual, structural, and manuscript. The skill prevents cascade problems where a change at one level unintentionally breaks elements at another level, whether a prose discovery reveals a structural flaw or a plot change requires extensive rewrites. Built for novelists working with Claude Code, Cursor or similar agents, it brings systematic change management to creative writing projects that typically become chaotic during revision passes. The diagnostic+assistive mode surfaces potential ripple effects before changes are committed, helping maintain narrative coherence across an entire book-length work.604installs17Story AnalysisStory Analysis is an agent skill that walks you through a fiction diagnostic after you finish a draft. Solo builders and indie authors who ship serial fiction, newsletters with narrative arcs, or game lore can invoke it instead of vague “does this work?” chat feedback. The skill splits work into Short Story mode for standalone pieces and Chapter mode for sections inside a larger work, each with focused evaluation blocks such as narrative foundation and character construction. You answer concrete questions—for example whether the core conflict fits in one sentence and whether scope matches word count—so revisions target structure and stakes rather than line edits only. It is evaluative, not generative: it does not rewrite prose; it surfaces strengths, weaknesses, and improvement opportunities aligned with whether the draft achieves its stated narrative goals. Pair it with your revision pass before beta readers, serialization, or launch.602installs18Story Idea Generatorstory-idea-generator is a generative fiction skill from jwynia/agent-skills (version 1.0, MIT license) that builds story concepts from emotional experience first rather than setting-first premises. The genre-first workflow ensures setting, characters, and plot serve the target emotional impact. Developers working on narrative games, interactive fiction, or creative side projects reach for story-idea-generator when starting a new story, strengthening a weak concept, or brainstorming genre-aligned ideas. The skill evaluates concepts against emotional goals and rejects hollow genre labels like sci-fi story without a defined reader experience.580installs19Task Decompositiontask-decomposition is a MIT-licensed utility skill at version 1.0 in jwynia/agent-skills that transforms overwhelming development work into manageable units. The diagnostic applies INVEST principles, vertical slices, spikes, and dependency analysis when users say a task is too big, cannot estimate, or needs epic breakdown. The skill integrates with github-agile for tracking decomposed work and runs in assistive mode on any project. Activate on keywords like decomposition, breakdown, estimate, scope, dependencies, or phrases like where do I start. Reach for task-decomposition when an agent or developer faces a monolithic feature that blocks estimation and needs independently completable chunks.568installs20Typescript Best Practicestypescript-best-practices is a jwynia/agent-skills reference with 494 installs on skills.sh and a rank of 21 that encodes idiomatic TypeScript conventions for agent-assisted development. The skill teaches agents to prefer precise types, safe narrowing, consistent module boundaries, and maintainable patterns while rejecting frequent anti-patterns like excessive any, improper async handling, and leaky abstractions. Developers reach for typescript-best-practices when reviewing agent output on React, Node, or full-stack TypeScript repos, or when setting project guardrails before large refactors. It functions as a standing style and quality checker rather than a test runner or formatter. Use it whenever TypeScript files are authored or edited by agents and long-term maintainability matters more than fastest first draft.555installs21Dialoguedialogue is a jwynia/agent-skills skill published on skills.sh with 476 installs and catalog rank 24 that guides agents toward natural, context-aware coding conversations. Instead of treating each message as an isolated command, dialogue keeps thread context, clarifies intent incrementally, and reduces brittle prompt-and-response patterns during implementation work. Developers reach for dialogue when agent sessions feel transactional, lose context between turns, or need a more collaborative conversational rhythm while editing code. The skill fits long-running Cursor or Claude Code sessions where back-and-forth discovery matters as much as final diffs.549installs22Fact Checkfact-check is version 1.0 MIT skill (author jwynia) that audits generated text for unsupported assertions, confabulations, and bad citations. It enforces a critical constraint: verification cannot be reliably combined with generation in a single pass, so agents run it AFTER content is produced. The methodology targets README sections, API docs, blog drafts, and commit-bound copy where plausible-sounding errors slip through. Developers reach for fact-check when accuracy matters more than speed—release notes, public docs, or research summaries where a wrong stat becomes a production incident. It behaves as an evaluative checker, not a writer, aligning with ship-phase review workflows.548installs23Story Sensestory-sense is a narrative craft skill from jwynia/agent-skills with 490 installs and rank 23 on skills.sh. It helps developers and coding agents inject story structure into product specifications, marketing copy, and system prompts so documents read with coherent arcs, meaningful stakes, and momentum instead of flat feature lists. Reach for story-sense when PRDs, landing page copy, or agent instructions feel technically correct but fail to persuade, orient, or sustain attention. The skill complements technical authoring by focusing on emotional through-lines and character motivation rather than code structure or API design.548installs24Story CollaboratorStory Collaborator is an agent skill that turns a solo builder into a creative partner for narrative-driven work. Whether crafting compelling user stories for a SaaS product, writing immersive onboarding flows, developing fictional product lore that makes marketing memorable, or building rich characters for a game or interactive experience, it keeps every element consistent. The skill remembers established rules, character traits, and plot points so you never have to repeat context. Builders use it when they need high-quality creative output faster than solo brainstorming allows but still want full control over direction and final decisions. It excels at long-form consistency where generic chat models tend to drift or contradict earlier details.525installs25Systemic WorldbuildingSystemic-worldbuilding is an agent skill that constructs rich, interdependent environments by modeling multiple overlapping systems such as economics, power structures, technology, culture, and ecology. Instead of generating isolated ideas, it produces coherent frameworks where every element has defined relationships, incentives, and consequences. Builders use it to create believable worlds that remain internally consistent even as the project grows across hundreds of prompts and iterations. The skill outputs structured world documents, rule matrices, character-system maps, and decision frameworks that agents can reference later, dramatically reducing drift and contradiction in long-form creative or simulation work.520installs26Pptx Generatorpptx-generator is a jwynia/agent-skills capability that lets Claude, Cursor, and similar agents assemble complete PowerPoint decks from short briefs, structured specs, or tabular data. The skill handles slide creation, layout selection, speaker notes, and file export so engineering teams can ship sales, training, or executive presentations without manual copy-paste from documents. Developers reach for pptx-generator when a feature launch, architecture review, or customer demo needs a .pptx deliverable alongside code. The workflow maps input requirements to slide outlines, applies consistent visual structure, and outputs an editable presentation file ready for stakeholders. Use pptx-generator when the artifact is a PowerPoint deck rather than markdown docs or PDF reports, and when agent assistance should cover both content drafting and deck structure in one pass.518installs27Skill Builderskill-builder is a jwynia/agent-skills package ranked #22 on skills.sh with 494 installs that guides developers through creating, packaging, and iterating reusable agent skills. The skill covers skill structure, metadata, triggers, and distribution so capabilities can be shared via Skillselion or installed by other agents. Developers reach for skill-builder when converting ad-hoc prompt patterns into maintainable, versioned skills with clear invocation boundaries. It fits platform engineers standardizing agent capability libraries across Claude Code, Cursor, and Codex environments. Rather than one-off instructions buried in project docs, skill-builder produces installable artifacts other team members and agents can discover and invoke consistently.518installs28Character ArcCharacter-arc is an agent skill that equips AI agents with rich, evolving character definitions including personality traits, emotional arcs, backstory, motivations, and growth triggers. It ensures that agents playing roles or interacting with users maintain believable consistency and demonstrate meaningful development over time instead of repeating the same patterns. Builders use it when creating narrative agents, customer personas, interactive storytelling tools, or any system where believable character behavior matters. The skill prevents generic responses and helps agents react differently as their internal state changes, making AI experiences feel more human and engaging.504installs29Research Workflowresearch-workflow is a structured agent methodology from jwynia/agent-skills for comprehensive research that spans planning, multi-query execution, credibility review, and synthesis into documented findings. Listed on skills.sh with 450 installs and rank 26, it activates when a single quick lookup is insufficient and a research report or auditable notes are expected. The workflow guides agents through repeated search-analyze-document cycles instead of one-shot answers, and it defers to lighter web-search skills when only a fast fact is needed. Developers reach for research-workflow when comparing libraries, investigating competitors, validating technical approaches, or gathering evidence for RFCs where source quality and cross-source agreement matter. The skill produces synthesized research output rather than code, so it fits early discovery work before validation or implementation begins. It is not meant for time-critical fact lookups or tasks that require no analysis beyond a single search result.479installs30Prose Styleprose-style is a jwynia/agent-skills writing governance skill for keeping generated prose aligned to a defined voice. It guides sentence rhythm, word choice, and tonal consistency across blogs, technical documentation, newsletters, and in-app copy so agent drafts do not read like generic LLM output. Developers and technical writers reach for prose-style when content ships alongside software—release notes, developer blogs, product copy, or long-form docs—and the team needs repeatable stylistic constraints rather than one-off rewrites. The skill acts as an editorial layer agents apply while drafting or revising text artifacts.476installs31Interactive Fictioninteractive-fiction is a diagnostic+assistive skill from jwynia/agent-skills (version 1.0, MIT license) that helps developers and narrative designers fix branching story problems. It triggers when choices feel meaningless, branching becomes unmanageable, player agency conflicts with authored plot, or interactive elements break narrative flow. The core principle is that agency and authorship coexist—player freedom and narrative coherence are balanced through structured diagnosis rather than generic writing prompts. Use interactive-fiction when building CYOA games, narrative prototypes, or playable fiction with state tracking and reader-facing presentation. The skill outputs actionable structural fixes for branching graphs, choice consequences, and flow pacing.472installs32Code Reviewcode-review is a Claude Code diagnostic skill (version 1.0) that provides structured code review guidance for catching defects and improving quality before merge. It activates on requests to review code, check for issues, run PR reviews, or evaluate maintainability, performance, and security across any programming language. The skill notes systematic review catches 60-90% of defects pre-production and integrates with github-agile for PR workflow context. Developers reach for code-review when they want evaluative, checklist-driven feedback instead of ad hoc nitpicks.469installs33Architecture Decisionarchitecture-decision is version 1.0 agent skill by jwynia that systematically evaluates architecture decisions, documents trade-offs, and selects appropriate patterns as Architecture Decision Records. The skill runs in diagnostic and evaluative mode when developers ask about ADRs, design pattern selection, technology choices, quality attributes, or technical debt implications. It integrates with system-design for ADR templates and works across any project stack. Reach for it before committing to structural changes—database selection, messaging patterns, auth models—so teams and agents retain context, rejected options, and rationale when code evolves or new contributors onboard.449installs34Shadcn Layoutsshadcn-layouts is version 1.0 MIT utility skill in jwynia/agent-skills that generates correct shadcn and Tailwind layouts by applying CSS mental models agents usually miss. It targets React and Next.js projects using shadcn/ui with Tailwind CSS v3 or v4, focusing on viewport-filling app shells, dashboard sidebars, marketing pages, flex column scroll regions, and grid collapse fixes rather than syntax errors alone. The skill inspects height constraint chains, recommends utilities such as h-full, h-screen, min-h-0, min-w-0, shrink-0, and grid classes at the correct parent levels, and checks missing component dependencies or Tailwind configuration issues. Trigger phrases include create a shadcn layout, fix layout issues, debug CSS height problems, make scrolling work, and Tailwind flex or grid failures. Developers reach for shadcn-layouts when generated UI collapses, nested scroll areas fail, or h-full elements shrink inside flex columns despite otherwise valid shadcn components. Install via npx playbooks add skill jwynia/agent-skills --skill shadcn-layouts or equivalent skills CLI flows.442installs35Outline Coachoutline-coach is version 1.0 assistive structural guidance skill by jwynia in the agent-skills repository. Licensed MIT and typed as diagnostic assistive mode, outline-coach asks questions, identifies structural issues, and suggests frameworks while explicitly never generating outline content—the writer or PM structures their own work. The skill supports fiction acts and chapters as well as PM epics and doc sections, helping define beats, dependencies, and cut lines before drafting or sprint planning. Developers and PMs reach for outline-coach when raw ideas need scoping discipline without an AI drafting the structure for them.434installs36Joke Engineeringjoke-engineering is a jwynia agent skill (version 1.0) that treats humor as engineerable connection systems rather than random creativity. In diagnostic and assistive mode, it analyzes why jokes fall flat—forced tone, weak misdirection, or punchlines that do not land—and recommends structural fixes using setup-punchline framing and revision passes tuned to audience and channel. Developers reach for joke-engineering when drafting conference talks, technical newsletters, release notes, or social posts where a deliberate comedic beat should clarify a point without undermining credibility.429installs37Key MomentsAgent skill for story structure validation: catalogs inciting events, progressive complications, midpoint shifts, crisis, climax, and resolution; ties each beat to character change and theme so authors test arc coherence early and avoid aimless middle acts.420installs38Game FacilitatorGame-facilitator turns agents into structured session hosts for RPGs and workshop games: track state, explain rules, prompt players, and keep pacing fair so interactive play runs smoothly without a human GM.412installs39Character Namingcharacter-naming is a fiction-focused generator skill in jwynia/agent-skills (version 1.0) that breaks LLM statistical name defaults with external entropy. It diagnoses median clustering (repeated Chen, Patel, Maya patterns), cast collision risks (Mark/Mike/Michael), and cultural incoherence, then generates names from curated cultural lists or fantasy phoneme presets with genuine randomization. Tools include cast-tracker for collision checking and JSON output for pipeline integration. Developers building games, visual novels, or NPC dialogue systems reach for character-naming when procedural cast rosters need phonologically consistent fantasy cultures or historically accurate period names. The skill runs a diagnostic flow before generation, enforcing context-first pools rather than raw LLM sampling. It complements game code skills by producing naming artifacts for quest scripts, localization CSVs, and character database seeds without trademark-like overlaps.409installs40Pwa Developmentpwa-development is a jwynia/agent-skills workflow for turning web apps into installable progressive web applications. The skill guides service worker setup, web app manifest configuration, responsive layouts, and navigation patterns that feel native on phones and desktops. Developers reach for pwa-development when a SaaS or mobile-first product needs offline resilience, home-screen installation, and app-shell UX without publishing to iOS or Android stores. It focuses on production PWA mechanics—caching strategies, manifest fields, and install prompts—rather than generic responsive CSS alone.408installs41Scene Sequencingscene-sequencing is a jwynia agent-skills workflow for arranging scenes in scripts, marketing campaigns, and interactive stories so agents maintain pacing, cause-effect continuity, and dramatic structure across multi-scene deliverables. The skill guides ordering decisions—what happens before and after each beat—so generated content reads coherently rather than as disconnected fragments. Developers reach for scene-sequencing when building narrative-driven products such as interactive fiction, game dialogue trees, video scripts, or multi-part launch campaigns that span several scenes. Agents apply it while drafting or refactoring scene outlines to enforce logical progression, tension curves, and payoff timing. Use scene-sequencing whenever multi-scene output needs deliberate ordering instead of one-shot paragraph generation.408installs42Musical Dnamusical-dna is a version 1.0 MIT-licensed agent skill from jwynia/agent-skills for Musical DNA Analysis. The skill extracts descriptive musical characteristics from any artist or band without using their name, building a technique-focused vocabulary of tempo, motifs, instrumentation, and mood for AI music generation, music description, or creative recombination. Developers reach for musical-dna when briefing audio partners or AI music tools for games, apps, trailers, or narrative worlds—replacing vague "sounds like [Artist]" references with specific, licensable sonic descriptors.405installs43Genre ConventionsSkill for mapping fiction genre conventions: hallmark tropes, pacing norms, ending contracts, voice expectations, and red-flag clichés across romance, thriller, fantasy, horror, and literary categories so outlines and drafts honor or subvert reader promises intentionally.403installs44Revisionrevision is a diagnostic agent skill (version 1.0) from jwynia/agent-skills—part of a 112-skill collection—that guides writers through systematic manuscript improvement after drafting. It defines three editing levels (developmental, line, copy) and six focused passes: structural, scene, character, dialogue, prose, and polish. Six revision states (R1–R6) diagnose overwhelm, author blindness, endless polishing, conflicting feedback, delete resistance, and bottom-up editing mistakes. A bundled revision-audit.ts Deno script tracks scene decisions and pass completion. The skill insists on top-down editing—fix structure before sentences—and includes darlings-folder tactics for painful cuts. Developers and technical writers reach for revision when a long draft needs pass-by-pass priorities instead of random fixes, especially when agents must recommend the next pass without rewriting the manuscript themselves.403installs45Flash FictionFlash-fiction enables agents to produce complete, punchy short stories with clear arcs and voice, optimized for marketing snippets, community posts, and editorial calendars that demand fresh narrative content quickly.402installs46Outline CollaboratorSkill for collaborating in Outline: structure collections, author and link wiki pages, and maintain shared team documentation so product knowledge stays searchable during build and operations.398installs47Economic Systemseconomic-systems is a version 1.0 Claude Code skill from jwynia/agent-skills that applies ten core principles of real economies to fictional worldbuilding. It generates currencies, trade networks, resource economies, and incentive structures so economic pressure drives plot and character motivation instead of static backdrop lore. Game developers and narrative designers reach for economic-systems when prototyping MMO economies, simulation games, or interactive fiction where scarcity, feedback loops, and market incentives must feel coherent. The skill operates in generative mode for worldbuilding domains rather than production code.397installs48Voice Analysisvoice-analysis is a Claude Code evaluative utility skill (version 1.0, MIT) for fiction and content writers who need style consistency. The skill analyzes source writing samples to fingerprint tone, rhythm, vocabulary, sentence structure, and recurring patterns, then outputs a structured voice guide agents can follow when drafting new copy. Developers and technical writers reach for voice-analysis when ghostwriting, scaling content teams, or training agents to match a brand voice without generic AI phrasing. The skill emphasizes capturing what makes the source distinctive—not keyword lists—so reproduction sounds like the original author.394installs49Gitea Workflowgitea-workflow is an agent skill from jwynia/agent-skills for self-hosted Gitea repositories. The skill guides branch creation, issue management, pull request workflows, merges, and release tagging through Gitea-native commands and APIs instead of defaulting to GitHub CLI patterns. Developers reach for gitea-workflow when their org runs Gitea on-premises or self-hosted and coding agents need correct issue, PR, and release workflows without assuming github.com remotes or gh CLI conventions.391installs50Oblique WorldbuildingOblique-worldbuilding applies lateral creative prompts to invent settings, cultures, and histories that feel surprising yet internally consistent. It is aimed at writers, GMs, and designers who need fresh worlds fast without defaulting to familiar fantasy or sci-fi templates.391installs51Settlement Designsettlement-design is a jwynia/agent-skills generative skill version 1.0 for designing realistic fictional cities and towns using ten core urban development principles and six parameter categories. Principles include geographic determinism, functional necessity, network emergence, layered development, power projection, resource constraint, cultural expression, adaptive reuse, disaster response, and spatial stratification. Parameter categories cover site selection, morphology, functional districts, infrastructure, socio-spatial layers, and symbolic elements so settlements feel lived-in rather than plot-driven. Developers and game writers reach for settlement-design when mapping city districts, faction territories, trade routes, and evolving urban history for RPGs, simulations, or collaborative fiction. The skill produces layered settlement documents with district functions, infrastructure maps, and historical accretion notes rather than production code artifacts.384installs52Draftingdrafting is a jwynia/agent-skills diagnostic skill at version 1.0 that helps break writer's block and produce first-pass drafts from briefs, outlines, or bullet notes when speed matters more than polish. The skill treats drafting as discovery rather than transcription, identifying why progress stalled and coaching execution until a complete rough draft exists for articles, memos, proposals, or creative pieces. Metadata marks it assistive with a fiction domain, but triggers cover any stalled writing once structure is defined. Developers reach for drafting when RFCs, blog posts, design docs, or stakeholder memos have outlines but no body copy. The workflow emphasizes momentum over perfection so downstream editing, review, and formatting skills can refine the text. drafting complements polish-oriented skills by filling the blank-page gap between planning and revision, asking what blocked execution and guiding incremental paragraphs until the draft exists for engineering communication or long-form documentation.383installs53Brainstormingbrainstorming is a version 1.1 agent skill from jwynia/agent-skills that structures AI-assisted ideation through two protocols: the five-phase Escape Velocity Protocol for breaking convergent thinking and the six-phase Seed Expansion Protocol for growing partial ideas. The skill diagnoses sessions across five states (B1–B5), separates solution form from required function, maps options on four axes (Who, When, Scale, Method), and audits idea orthogonality before evaluation. A bundled Deno/TypeScript script, constraint-entropy.ts, injects random constraints from five categories—actors, resources, combinations, inversions, and domains—to force exploration humans and LLMs typically skip. Session outputs persist as dated markdown files covering enumerated defaults, axis maps, entropy prompts, and shortlisted concepts. Developers reach for brainstorming when early product or feature discussions keep recycling the same options, or when they have a seed concept but need adjacent moves before writing prototypes, PRDs, or implementation plans.380installs54Security Scansecurity-scan is a skill from jwynia/agent-skills that orchestrates structured security scanning across application code and configuration files before release. The agent follows a defined scan workflow to identify dependency CVEs, insecure defaults, exposed secrets patterns, and compliance gaps rather than ad-hoc grep searches. Developers invoke security-scan when preparing a release candidate, onboarding a inherited repo, or responding to audit requests. Output is organized findings with severity and remediation guidance suitable for ticket creation. The skill complements manual code review by systematically covering common OWASP and misconfiguration categories across the repository tree.379installs55Shared Worldshared-world is a MIT-licensed agent skill (version 1.0) from jwynia/agent-skills that treats a fiction universe like team memory for software: organized entries, cross-references, and navigation through characters, locations, history, factions, rules, culture, artifacts, species, and meta. Each entry records canon status across five states—Established, Proposed, Deprecated, Contradicted, and Speculative—plus sources, contributors, relationships, and a changelog. Bundled tooling initializes a bible, adds templated entries, rebuilds indexes, scans for contradictions, validates links, and exports a single-file reference. A developer reaches for shared-world when spinning up a long-running shared universe, onboarding new contributors, or running multi-agent narrative pipelines that must not break continuity across branches.379installs56Story ZoomApplies a story-zoom lens—wide mythic stakes down to single sensory beats—to sharpen premises, scene lists, and character arcs for fiction, games, or serialized content before committing to full manuscripts or production pipelines.378installs57Reverse Outlinerreverse-outliner is a jwynia/agent-skills fiction utility that extracts story architecture from finished prose through a Deno pipeline. Version 1.0 orchestrates segment-book.ts, analyze-scene-batch.ts, and generate-outline.ts via reverse-outline.ts to produce outline.md plus intermediate JSON artifacts. The workflow detects chapters and scene breaks, applies Goal/Conflict/Disaster analysis, maps genre-specific key moments, traces character arcs, and synthesizes markdown outlines at configurable depth. Developers and technical authors reach for reverse-outliner when studying craft from masters, building classroom materials, or finding weak transitions and redundant sections in long-form drafts. Run deno run --allow-read --allow-write scripts/reverse-outline.ts book.txt with optional --genre and --output flags to emit a reverse-outlines directory.377installs58Namingnaming is a diagnostic and generative skill from jwynia/agent-skills, part of a collection of 112 reusable Agent Skills. Version 1.0, it diagnoses naming problems across brands, products, characters, places, and titles by identifying naming states—feels wrong, family mismatch, forgettable, wrong signals, or practical failures—and analyzing four layers: sound, meaning, cultural resonance, and functional fit. Professional branding uses a sequential phased process; quick naming applies diagnostic states directly for characters, places, or simple products. The parent repo documents agentskills.io naming rules: 1–64 lowercase hyphenated characters matching the SKILL.md name field. Developers reach for naming when a product, skill, repo, or feature name feels off and they need structured diagnosis instead of random brainstorming.376installs59Xlsx Generatorxlsx-generator is a Claude Code skill for producing production-quality Excel workbooks programmatically. It guides agents to create multi-sheet XLSX files with formulas, cell formatting, and embedded charts suitable for finance models, operational reports, and stakeholder exports. Developers reach for it when an application or agent pipeline must emit spreadsheet deliverables instead of CSV dumps or PDF summaries. The skill covers layout conventions, formula patterns, and chart placement so output opens cleanly in Excel and LibreOffice. Use it for automated reporting endpoints, batch export jobs, or agent-generated analysis packages.376installs60Governance SystemsProvides patterns for building governance systems around AI agents, including policy enforcement, human review gates, audit trails, and organizational rule encoding. Helps teams structure safer autonomous agents with clear accountability, permission boundaries, and operational controls before deployment.374installs61Ebook Analysisebook-analysis in jwynia/agent-skills performs non-fiction knowledge extraction from ebooks and long PDFs with mandatory citation traceability to exact sources. The skill supports two complementary extraction modes: Concept Extraction classifies ideas from principle down to tactic, and Entity Extraction captures named studies, researchers, frameworks, and anecdotes that persist across books. The core rule is to extract less with full provenance rather than more without it, making outputs suitable for reviews, study guides, and derivative content planning. Developers reach for ebook-analysis when summarizing technical books, building reading notes with quotable citations, or comparing frameworks across multiple titles. Agents produce chapter-level summaries, argument maps, and insight lists tied to page or section references. The workflow favors traceable extractions over bulk paraphrase, which keeps downstream publishing and documentation auditable.372installs62Book Marketingbook-marketing is a jwynia/agent-skills fiction-application utility that helps translate finished manuscripts into converting marketing materials. The skill runs diagnostic states for vague positioning, genre mismatch, and platform-specific copy problems, then generates four deliverable types: back-of-book blurbs, Amazon HTML descriptions, social taglines and hooks, and 150-200 word agent query pitches. Developers writing technical books reach for book-marketing when retailer metadata, comp titles, promo calendars, or newsletter and social channel copy must align with genre conventions. The workflow selects deliverables based on destination—physical blurbs for print, HTML for Amazon listings, short hooks for ads, and query letters for agent submissions—rather than producing one generic description.371installs63Chapter Drafterchapter-drafter helps agents produce complete book chapters from outlines or briefs, maintaining tone, scene logic, and narrative momentum across long-form creative or informational writing projects.370installs64Cliche Transcendencecliche-transcendence is a version 1.0 diagnostic-plus-generative agent skill from jwynia/agent-skills that helps developers writing fiction, game narrative, or long-form content replace predictable story beats with fresher alternatives. The skill runs an eight-step CTF process, applies the Orthogonality Principle across Form, Knowledge, Goal, and Role axes, and tests whether elements collide with the protagonist instead of conveniently serving the plot. Metadata marks domain fiction with MIT license, and outputs can persist to project files between sessions. Reach for cliche-transcendence when drafting or revising scenes that feel generic, receive 'seen this before' feedback, or orbit the protagonist too obediently—common in game writing and technical storytelling side projects.369installs65Belief SystemsBelief-systems teaches agents to represent convictions, confidence, and revision rules so multi-turn reasoning, role-play, and deliberation stay internally coherent as new information arrives.368installs66Gentle Teachinggentle-teaching is a version 1.0 Claude Code utility skill from jwynia/agent-skills that translates gentle-parenting principles into adult education boundaries for AI-assisted learning. It prioritizes process over solutions, empathy, developmental awareness, and clear limits so learners build independence instead of agent dependence. Developers reach for gentle-teaching when onboarding juniors, explaining unfamiliar frameworks, or helping teammates who struggle with dense technical concepts. The skill adapts pacing with formative comprehension checks rather than overwhelming detail dumps or one-line replies.368installs67EndingsSpecializes agents in writing powerful endings: evaluating setup payoffs, choosing closure types, weaving callbacks, and avoiding weak or generic finales for fiction, essays, and branded storytelling.364installs68Language Evolutionlanguage-evolution in jwynia/agent-skills (version 1.0, MIT, type generator) helps developers and technical writers model how languages, dialects, and domain vocabularies change over time to inform naming conventions, localization choices, inclusive product copy, and fictional worldbuilding. The skill applies historical continuity so new terms branch credibly from parent languages, reflect cultural contact, and document sound shifts, borrowing, and semantic drift across generations and geographies. Developers reach for language-evolution when product glossaries, fantasy settings, or localization style guides need more than ad hoc word lists—especially when terminology must stay internally consistent across releases, regions, or narrative timelines. The workflow moves from proto-language roots through dialect splits, contact effects, and modern register choices, producing structured vocabulary and evolution notes rather than one-off neologisms. With metadata domain worldbuilding and generative mode, language-evolution fits teams shipping games, speculative fiction tooling, or inclusive UX copy where language credibility affects user trust and immersion quality.363installs69Researchresearch is an agent skill for structured information gathering when exploring markets, technologies, or problem spaces before committing to a build plan. It produces sourced comparisons, concise summaries, and citable references rather than jumping straight into code. Developers reach for research when they need disciplined discovery—competitive landscape, technology options, or domain context—before validation or implementation work begins. The skill emphasizes traceable sources and decision-ready summaries over open-ended brainstorming.361installs70Revealjs Presenterrevealjs-presenter is a jwynia/agent-skills generator skill at version 1.0 that produces single-file Reveal.js HTML presentations with embedded CSS and reveal.js 4.6.1 CDN assets. It configures 1920×1080 slides with fade transitions, documents slide types from title and impact to code and comparison layouts, and supports RevealJS fragments for progressive disclosure plus speaker notes opened with the S key. Three bundled theme specifications—bold, corporate, and delft—define typography, word limits, and content rhythm before generation begins. Developers reach for revealjs-presenter when converting outlines into conference talks, engineering demos, pitch decks, or training decks without building a separate slide toolchain. Output is a portable HTML file served over HTTP for presenter view, keyboard navigation, and hash-based deep links between slides. The skill warns against fighting RevealJS centering with conflicting flexbox, recommends professional fade transitions, and includes footer branding patterns plus accessibility-minded typography using Inter, Space Grotesk, and JetBrains Mono stacks.361installs71Context Retrospectivecontext-retrospective is a jwynia/agent-skills utility (version 1.0, evaluative mode) that reviews agent-user transcripts to improve context networks and agent instructions. It examines six analysis dimensions—including knowledge gaps, navigation inefficiency, boundary violations, and guidance failures—and maps findings to missing nodes, broken relationships, and unclear rules. Output includes prioritized recommendations rated critical through low plus edit templates for knowledge gaps, navigation issues, and guidance assessments. The 10.5 KB SKILL.md defines a retrospective process with preparation, chronological transcript review, and concrete action items. Developers run context-retrospective after significant agent sessions, incidents, or writing sprints when agents repeatedly rediscover information or violate document boundaries.360installs72World FatesDesigns world-fate frameworks—omens, prophecies, inevitable twists, and player agency tradeoffs—so games and stories share predictable yet surprising outcome logic instead of ad hoc plot twists that break established rules.356installs73Sensitivity Checksensitivity-check is an evaluative agent skill in jwynia/agent-skills that reviews copy, UX flows, and AI-generated outputs for representation harm, biased stereotypes, privacy pitfalls, regulatory wording issues, and brand-risk language before shipping. It provides context for safer deployment of user-facing strings, onboarding flows, marketing pages, and agent responses that could expose teams to reputational or compliance problems. The skill sits alongside claim-investigation and fact-check in the repository and explicitly supports evaluating representation claims during investigative workflows. Product engineers, frontend developers, and technical writers reach for sensitivity-check during pre-release content review, UX audits, or agent output QA—not when implementing backend logic or writing unit tests. It complements design review by focusing on harm flagging and inclusive language rather than visual layout or accessibility mechanics.354installs74Task Breakdowntask-breakdown is version 1.0 utility skill by jwynia that decomposes overwhelming or vague goals into ordered, estimable tasks with dependencies, acceptance hints, and right-sized chunks before implementation or sprint planning. Originally designed with ADHD and autism-aware decomposition strategies, the skill provides external scaffolding for executive function support rather than motivation lectures. Developers reach for task-breakdown when a goal feels too large to start, sprint backlog items need clearer slices, or an agent must translate ambiguous requirements into actionable steps with explicit ordering. The skill applies when tasks feel paralyzing or when someone cannot determine a concrete first move. It fits engineering leads, individual contributors, and agent workflows that need structured work breakdown before writing code or opening tickets.354installs75Lyric DiagnosticDiagnostic skill for song lyrics: evaluates rhyme schemes, rhythm, clichés, story clarity, and emotional payoff, then proposes concrete rewrites so verses feel intentional and performance-ready.350installs76Memetic DepthAgent skill for memetic narrative depth: traces how concepts replicate and evolve across chapters, designs recurring symbols with escalating meaning, flags shallow allegory, and helps authors build stories that reward rereading through layered cultural and thematic resonance.350installs77Scraper Builderscraper-builder is a skill from jwynia/agent-skills focused on production-grade web scraping architecture rather than one-off curl commands. It guides agents through DOM selector strategy, paginated crawl loops, respectful rate limiting, durable storage sinks, and retry-aware error handling so extracted data feeds research pipelines, pricing monitors, or downstream agent tools reliably. Developers invoke it when targeting semi-structured public pages, catalog listings, or paginated APIs disguised as HTML. The skill emphasizes resilience against layout drift, backoff policies, and idempotent writes instead of fragile single-page scripts. Use scraper-builder when building maintainable extractors that must run repeatedly without manual intervention.350installs78Good Thinkinggood-thinking is a skill from jwynia/agent-skills that front-loads structured reasoning into agent sessions before any build work starts. It prompts clarification of goals, explicit assumption listing, tradeoff analysis, and decision framing so later implementation steps stay aligned with intent. Developers invoke it when a feature request is ambiguous, when multiple architectural paths exist, or when an agent risks jumping straight to code without validating constraints. The skill acts as a cognitive checkpoint rather than a code generator, producing clearer problem statements and decision records that downstream skills can execute against. Use good-thinking at session start for complex features, refactors with competing approaches, or any task where unstated assumptions could derail delivery.349installs79Moral ParallaxTrains agents to apply moral parallax: juxtaposing credible ethical viewpoints on the same choice, clarifying conflicts, and producing nuanced rationale before policy, product, or autonomous action design.346installs80Sleep Storysleep-story is an agent skill from jwynia/agent-skills version 1.0 classified as a generative fiction tool for bedtime content. It guides agents to write verbal lullabies disguised as narratives that occupy the mind just enough to prevent racing thoughts while fading as attention wavers. The skill emphasizes gentle cognitive load, low-stakes arcs, soft sensory imagery, and pacing that slows naturally toward rest. It targets bedtime podcasts, meditation apps, calming newsletters, and any product where the goal is relaxation rather than plot tension. Use sleep-story when you need structured prompts and narrative constraints for sleep-oriented audio scripts or serialized wellness content rather than action-driven fiction.346installs81Claim Investigationclaim-investigation is version 1.0 of a diagnostic research skill in jwynia/agent-skills for systematically investigating social media assertions, viral posts, and document claims before drawing conclusions. It decomposes complex statements into atomic components classified as entity, event, state, process, causation, or narrative, resolves vague references to specific names and dates, and verifies foundational facts using prioritized primary sources such as official records, court documents, and cross-checked news reports. The skill separates confirmed facts from disputed items and narrative framing, reporting confidence levels and alternative explanations in a source-backed synthesis. Developers reach for claim-investigation when triaging user-reported claims, evaluating technical rumors, or scoping what evidence exists before product decisions. It pairs with sibling skills fact-check and sensitivity-check in the repository's seven-skill Research category.345installs82Media AdaptationSkill for adapting fiction across media: identifies what survives medium changes, converts internal monologue to visual or audio storytelling, restructures acts for episodic or runtime limits, and outputs pitch-ready briefs for screen, audio drama, or interactive formats.345installs83Agile Workflowagile-workflow in jwynia/agent-skills helps engineering teams run agile ceremonies with agent assistance for tickets, acceptance criteria, and incremental delivery tracking. The skill supports sprint planning sessions that break epics into scoped stories, backlog grooming that refines priorities and definitions of ready, daily standup summaries capturing blockers and progress, and retrospectives that document improvements and action items. Agents draft user stories with testable acceptance criteria, estimate complexity, and track sprint commitments without replacing human prioritization decisions. Developers reach for agile-workflow when a coding agent should participate in ceremony prep, convert meeting notes into structured tickets, or maintain lightweight sprint documentation alongside implementation work. It complements general-purpose project skills by focusing on recurring Scrum-style loops rather than one-off specifications. Triggers include sprint planning, backlog grooming, standup notes, retrospectives, and acceptance criteria writing.344installs84ConlangGuides Claude through constructed-language creation for fiction: sound inventories, word-formation patterns, sentence structure, orthography options, sample translations, and a maintainable conlang bible authors reuse while drafting and revising speculative manuscripts.340installs85Dna Extractiondna-extraction is a Claude Code skill for developers configuring AI agents who need consistent voice, architecture decisions, or brand behavior across long-running projects. The skill analyzes sample code, copy, or design artifacts to distill recurring style patterns, stated values, hard constraints, and behavioral norms into a compact project DNA profile. Developers reach for dna-extraction when agents drift from established conventions between sessions or when onboarding a new agent context requires encoding tacit project rules that never made it into formal docs. The output feeds agent system prompts, skill libraries, and codegen guardrails so generated code and copy match the project's established character.340installs86Adaptation Synthesisadaptation-synthesis is a Claude Code skill from jwynia/agent-skills that helps developers consolidate feedback, experiment results, and evolving requirements into actionable adaptations for agents or early prototypes. The skill supports structured iteration when behavior must change before committing to a full implementation path. Reach for adaptation-synthesis after user tests, agent eval runs, or stakeholder feedback sessions that surface conflicting signals. It fits AI engineers and product-minded developers refining agent prompts, tool use, or prototype scope without prematurely locking architecture. Use it when multiple experiment threads need synthesis into a coherent next iteration plan rather than ad hoc tweaks.334installs87Table ToneHelps GMs and narrative agents set, monitor, and preserve tabletop session tone—genre, tension, humor, and pacing—so play feels intentional and consistent from scene to scene.334installs88Multi Order EvolutionSkill for modeling multi-level narrative evolution: links individual decisions to group dynamics and cultural drift, simulates second- and third-order consequences across story eras, and helps authors discover plausible futures, factions, and theme arcs before outlining.333installs89Statistical Distancestatistical-distance equips agents to pick and apply statistical distance functions—KL, Wasserstein, cosine-related measures, and related tests—for comparing distributions in code. It supports backend analytics, ML features, and QA of similarity or drift detection implementations.333installs90Metabolic Culturesmetabolic-cultures is a creative writing skill that applies biological-metaphor frameworks to develop layered fictional cultures, paradoxical societies, and narrative worlds for long-form content and storytelling projects.332installs91Identity DenialThe identity-denial skill from jwynia/agent-skills helps secure Claude agent workflows by preventing inappropriate revelation of identity-related information. It is aimed at teams shipping agent features who need explicit guardrails so automated assistants do not expose sensitive attribution, credentials, or policy-violating identity details during real usage.331installs92Positional RevelationThe positional-revelation skill from jwynia/agent-skills provides patterns for managing how coding agents reveal positional or situational context during execution. It helps builders ship safer agent systems that disclose only task-relevant state instead of dumping broad workspace, path, or environment details that could increase security and privacy risk.331installs93Context Networkcontext-network is a context-architecture skill for developers running long agent sessions on complex codebases where flat prompts lose thread. It instructs agents to represent facts, files, decisions, and tasks as linked nodes so retrieval pulls coherent background instead of stale or conflicting snippets. The pattern reduces contradictory answers when multiple decisions, specs, and task states evolve over weeks. Use context-network when a project spans many files and milestones and the agent must maintain a durable, queryable map of what was decided and why.330installs94Perspectival ConstellationPerspectival-constellation teaches Claude agents to analyze topics through a coordinated set of viewpoints, reducing blind spots and producing richer discovery outputs before scope or build decisions are made.328installs95Underdog UnitUnderdog-unit guides agents to validate scope through a contrarian underdog lens, highlighting overlooked constraints, asymmetric bets, and clearer tradeoffs before implementation begins.328installs96List Builderlist-builder is a jwynia agent-skills utility that transforms unstructured notes or requests into prioritized, structured lists developers can execute against. It supports product backlogs, content plans, feature scopes, and launch checklists by ranking items, grouping related work, and clarifying sequencing before coding or publishing begins. Developers reach for list-builder when brainstorming produced too many ideas and a single ordered artifact is needed for sprint planning, content calendars, or release prep. The skill does not implement features—it produces planning lists that pair well with PRD or implementation skills downstream. It suits SaaS roadmaps, agent task queues, and content pipelines where priority and grouping reduce thrash. Prompt with constraints, deadlines, and audience so rankings reflect real tradeoffs rather than generic bullet dumps.326installs97Paradox Fablesparadox-fables is a jwynia/agent-skills generative skill at version 1.0 that helps writers produce short fables embodying paradoxical wisdom without collapsing into simple morals. Unlike traditional fables that resolve cleanly, paradox fables maintain tension so readers absorb lessons through story rather than direct argument. The skill metadata marks it as a generator in the fiction domain and triggers when exploring irresolvable tensions, bypassing analytical defenses, or crafting teaching stories for publication. Reach for paradox-fables when drafting newsletter allegories, blog parables, or social posts that illustrate contradictions and moral dilemmas. Skip it when you need technical documentation, API references, or factual research summaries instead of narrative content.323installs98Abstract Strategyabstract-strategy is a game-design agent skill (version 1.0, MIT) for creating abstract strategy games defined by perfect information, zero randomness, and deep strategic play. It provides frameworks for ideation, mechanic design, and balance evaluation when brainstorming board games or digital equivalents. Developers reach for it when exploring move spaces, win conditions, and fairness before coding prototypes or playtesting sessions. Keywords in the manifest include board game, mechanics, strategy, and balance, signaling a generative workflow rather than engine implementation.321installs99Product Analysisproduct-analysis is version 1.0 of jwynia/agent-skills' competitive diagnostic skill. It detects which of eight analysis states (CPA0 through CPA7) a project occupies—from no analysis through assumed competition, features without taxonomy, missing prevalence data, lacking personas, unmapped features, deferred decisions, to analysis complete—then routes to one of six interconnected frameworks such as Competitive Niche Boundary, Feature Taxonomy, Persona Construction, and Build/Buy/Partner analysis. The skill focuses on jobs-to-be-done and evidence-based personas rather than feature-counting bias. Outputs include validated competitor rosters, canonical feature taxonomies, table-stakes versus differentiator matrices, prevalence heatmaps, persona profiles, and strategic build-buy-partner recommendations. Developers reach for product-analysis when evaluating market entry, building feature comparisons, or deciding whether to enter a product category before writing code.321installs100Speech Adaptationspeech-adaptation is a generative utility skill (version 1.0, MIT license) from jwynia/agent-skills that transforms comprehensive written content into purposeful spoken guidance. Speech requires roughly 3-5x compression while maintaining functional value, so the skill restructures dense prose into listenable scripts suited for podcasts, presentations, voice assistants, and audio courses. Developers reach for speech-adaptation when converting blog posts, tutorials, or internal docs into narration, optimizing copy for listening comprehension, or adapting dialogue voice and dialect per character while preserving identity across scenes. The skill applies domain-specific writing rules for pacing, emphasis, and oral clarity rather than simple summarization. It fits content pipelines where written technical material must become high-quality audio without losing actionable guidance.319installs101Competency Buildercompetency-builder is a diagnostic agent skill (version 1.0) that helps developers and engineering leads design competency frameworks where capabilities are observable behaviors, not checkbox completion. The skill inspects existing training materials, competency statements, and failure modes, then maps symptoms to eleven diagnostic states (CF0 through CF10) such as content-first traps, vague competencies, or missing feedback loops. It outputs competency definitions, scenario designs, audience-layered progression trees, verification rubrics, and feedback-loop plans using bundled templates for competencies, scenarios, progression, and gap tracking. Developers reach for competency-builder when onboarding programs, internal certifications, or agent-training curricula need structured evidence models before writing course content or shipping learning tooling.317installs102Agile Coordinatoragile-coordinator is a multi-agent orchestrator skill (version 1.0) from jwynia/agent-skills that coordinates worker agents executing agile-workflow on ready backlog tasks. It discovers tasks from a context-network backlog, plans sequential or parallel execution with up to N workers (default 2), spawns workers via Claude Code's Task tool, merges feature branches one at a time, runs npm build and test verification, and persists progress back to backlog epic files across seven workflow phases. Developers reach for agile-coordinator when multiple groomed tasks need autonomous batch execution with conflict-safe merges and post-merge verification instead of hand-merging each branch.313installs103Npm Packagenpm-package is an agent skill from jwynia/agent-skills that teaches a Bun-first workflow for creating and publishing npm-compatible TypeScript libraries. The toolchain pairs Bun as runtime and package manager, Bunup for ESM and CJS bundle generation with declaration files, strict TypeScript with module nodenext, Biome v2 plus ESLint for linting, Vitest for unit tests, and Changesets for versioning and changelog management. Developers reach for npm-package when scaffolding a new library, fixing CJS and ESM interop or package.exports map issues, configuring conditional exports and .d.ts generation, or preparing CI-ready npm publish pipelines. The skill documents when to create packages from scratch versus migrating existing code, resolving default and named export compatibility across module systems, and reviewing package.json fields before registry release. Keywords include npm, bun, bunup, esm, cjs, vitest, biome, and changesets.310installs104Blind Spot Detectiveblind-spot-detective is an MIT-licensed Claude skill from jwynia/agent-skills at version 1.0 with evaluative diagnostic mode for non-fiction writing. The skill distinguishes blind spots—inherent limitations of an approach—from blank spots—addressable gaps—and applies frameworks for finding omissions and testing assumptions. Metadata classifies the domain as writing with type diagnostic. Developers reach for blind-spot-detective before finalizing technical essays, guides, or explanatory posts when feedback feels incomplete. The skill is not a code diff reviewer; it targets prose completeness and assumption coverage in non-fiction work.300installs105Non Fiction Revisionnon-fiction-revision is a diagnostic writing skill (version 1.0, MIT license) for improving non-fiction drafts without changing core arguments. Author jwynia built it for business, self-help, academic, popular science, memoir, and technical documentation where thesis, structure, evidence, and reader confusion must be addressed simultaneously. Developers writing long-form docs, internal reports, or published articles invoke it when arguments feel weak, evidence is outdated, readers report confusion, or book-level structure has problems. The skill operates in diagnostic plus assistive mode, separating structural issues from line-level clarity so revisions target the right layer of the manuscript or document.298installs106Devcontainerdevcontainer is an agent skill from jwynia/agent-skills for configuring VS Code/Cursor devcontainers so AI-assisted development runs inside a reproducible Docker-based environment. It helps developers define devcontainer.json, Dockerfile references, feature packs, and post-create commands so every contributor gets the same language runtimes, CLIs, and extensions. Teams reach for it when onboarding is slow, local dependency drift breaks agent workflows, or remote-container development must mirror CI. The skill focuses on environment parity for coding agents, not production deployment pipelines.297installs107Gitea Coordinatorgitea-coordinator is a Claude Code orchestrator skill at version 1.0 that coordinates multiple worker agents to implement groomed backlog tasks in self-hosted Gitea repositories. The skill executes six workflow phases—discovery, planning, execution, merging, verification, and summary—reading ready tasks from a context-network backlog, spawning workers through Claude Code's Task tool, and delegating code changes to the gitea-workflow skill. Workers write progress to .coordinator/workers/{worker-id}/progress.json while the coordinator queues pull requests, merges them sequentially with the Gitea Tea CLI, deletes feature branches, and runs npm build and npm test verification after all merges complete. Command-line flags including --dry-run, --parallel, --max-workers (default 2), --autonomous, and --supervised control batch execution without allowing parallel merges. Developers reach for gitea-coordinator when several groomed backlog items need autonomous implementation on Gitea with coordinated merges and post-merge regression checks.297installs108Mastra Honomastra-hono is an agent-building skill for scaffolding Mastra AI applications exposed through Hono HTTP routes. It helps developers wire agent tools, memory stores, and workflow orchestration into a lightweight, edge-ready API surface suitable for serverless or minimal Node deployments. Teams reach for it when starting a Mastra project that needs REST or fetch-handler endpoints instead of a heavy full-stack framework. The skill emphasizes route structure, agent invocation patterns, and Mastra-Hono integration rather than generic Express or Next.js setups.296installs109Github Agilegithub-agile is a diagnostic agent skill (version 1.0) from jwynia/agent-skills that helps developers establish and maintain lightweight agile execution on GitHub. The skill maps eight workflow states—from missing GitHub CLI through healthy delivery—and prescribes interventions using gh commands, four bundled Deno scripts (gh-verify, gh-init-project, gh-audit, gh-sync-context), and context-network files such as status.md and decisions.md. It initializes labels across three schemes, issue templates for features, bugs, and tasks, branch protection, and PR templates, then enforces feature-branch naming, MoSCoW backlog grooming, and milestone sync ceremonies. Reach for github-agile when a repository has chaotic issues, stale PRs, direct-to-main commits, or a context network that no longer matches GitHub reality.293installs110Technology Impacttechnology-impact is a jwynia/agent-skills utility at version 1.0 that systematically analyzes societal and organizational consequences of technologies using McLuhan's Tetrad of Media Effects. The framework examines four simultaneous dimensions—what a technology enhances, obsoletes, retrieves, and reverses—to reveal non-obvious adoption risks that roadmap spreadsheets miss. Developers and tech leads invoke technology-impact when evaluating new frameworks, planning platform migrations, or drafting technology policy before architecture review boards sign off. The skill's evaluative mode targets worldbuilding and adoption planning workflows where second-order effects on teams, cost, and delivery timelines matter as much as feature checklists. Reach for technology-impact during vendor bake-offs, build-vs-buy decisions, or platform shifts when stakeholders need structured consequence mapping instead of hype-driven slide decks. Outputs are tetrad worksheets and impact narratives engineers attach to RFCs, ADRs, and adoption proposals so trade-offs stay visible through implementation.286installs111Npx Clinpx-cli is a CLI toolchain skill in jwynia/agent-skills—part of a 112-skill collection—that teaches building and publishing npx-executable command-line tools with Bun as the primary runtime while producing Node.js-compatible npm binaries. The SKILL.md covers project scaffolding, citty argument parsing with runMain() and sub-commands, picocolors terminal UX, strict TypeScript with module nodenext, Biome v2 formatting, ESLint with typescript-eslint, Vitest testing, Bunup dual-entry bundling for library and CLI outputs, Changesets versioning, and npm publish --provenance releases. Key rules enforce files whitelist packaging, Node shebangs on published bins, and separating thin CLI wiring from importable core modules. Developers reach for npx-cli when creating a new agent-invokable CLI, adding a bin entry to an existing library, or publishing typed tools consumable via npx without manual shell choreography. Install with npx skills add jwynia/agent-skills --skill npx-cli from the skills/tech/development/tooling path.283installs112Skill Integratorskill-integrator is a meta utility skill (version 1.0) in jwynia/agent-skills that closes the gap between installed skills and agent awareness. It runs a Deno pipeline—analyze-project.ts, scan-skills.ts, and generate-guidance.ts—to detect project type, inventory .claude/skills, score relevance, and emit a marked guidance section for CLAUDE.md or AGENTS.md. Four diagnostic states (SI0–SI3) cover first-time setup, stale sections after skill changes, single-skill additions, and misfit trigger categories. Developers reach for skill-integrator after installing skills, bootstrapping a new repo, or when agents never invoke available tools despite a full skills directory. Single-skill mode adds one entry; full mode regenerates the entire <!-- Generated by skill-integrator --> block with project-specific trigger phrasing instead of echoing SKILL.md descriptions verbatim.283installs113Web Search Braveweb-search-brave is an agent skill that wires Brave Search into coding workflows through a Deno CLI script against api.search.brave.com. After setting BRAVE_API_KEY, developers run searches that return up to 20 results with titles, URLs, descriptions, and optional extra snippets, using flags for country codes, language, safe-search, freshness windows (pd, pw, pm, py), pagination offset, and JSON output. The skill targets privacy-respecting retrieval when built-in search lacks localization or time controls, such as verifying release notes, regional docs, or news during feature work. It complements generic web-search and Tavily-style ranked search when a Brave API key is already provisioned. Prerequisites include Deno, network access, and a Brave Search API plan; extra snippets may require a paid tier. Version 1.0 ships in the jwynia/agent-skills research tools bundle.283installs114Secrets Scansecrets-scan is version 1.0 MIT utility skill by jwynia that performs deep detection of hardcoded credentials and sensitive data in source code. The SKILL.md documents CLI-style invocations such as `/secrets-scan`, `/secrets-scan --scope src/`, `/secrets-scan --entropy` for high-entropy strings, and `/secrets-scan --git-history` to inspect commits. It classifies high-confidence secret patterns including API keys, passwords, and tokens, and supports scoped directory scans for monorepos. Developers run it before pull request merge, release tagging, or after refactors that touch config files and `.env` examples. The evaluative mode fits CI pre-commit hooks and agent-assisted reviews where humans want a structured pass beyond ad hoc regex searches. It reduces incident risk from accidental credential commits without replacing dedicated secret managers or vault rotation workflows.275installs115Dependency Scandependency-scan is version 1.0 evaluative utility skill in jwynia/agent-skills for detecting known vulnerabilities in project dependencies. The skill scans detected package managers or targets specific ecosystems with flags like /dependency-scan --npm, /dependency-scan --pip, and /dependency-scan --fix for auto-remediation where possible. It inspects manifests and lockfiles for CVEs, outdated packages, license conflicts, and transitive dependency chains that manual review often misses. Developers reach for dependency-scan during pre-release security checks, dependency upgrade planning, and CI hardening across Node.js, Python, Rust, and additional ecosystems. The MIT-licensed skill runs as an agent-invoked audit rather than a hosted scanner, fitting local repo workflows before merge or deploy.264installs116Media Meta Analysismedia-meta-analysis is a Data Science & ML research utility skill from jwynia/agent-skills at version 1.0. The skill operates on collections of already-analyzed media content, not single articles, to reveal cross-source patterns, conceptual networks, and emergent themes invisible in isolated reviews. It supports comparing themes, sentiment patterns, and effect sizes across a corpus before committing to production content or research claims. Developers and researchers reach for media-meta-analysis when they hold multiple prior media analyses and need a structured meta-analysis layer that connects findings across sources instead of re-reading each report manually.262installs117Config Scanconfig-scan is a version 1.0 MIT-licensed utility skill by jwynia that performs evaluative security review of configuration files and infrastructure-as-code. Developers invoke /config-scan to scan all config files or narrow scope with --docker, --k8s, --terraform, or --env flags for container definitions, Kubernetes manifests, Terraform modules, and environment settings. The skill targets secrets exposure, insecure defaults, and environment-specific risks before merge or production deploy. It fits pre-release checklists and periodic security audits where misconfigured ports, overly permissive IAM, or committed credentials would otherwise slip past code review. Output is review-oriented rather than auto-remediation, keeping humans in the loop for fixes while giving agents a consistent audit command across polyglot repos mixing app config and IaC.256installs118Agent Bootstrapagent-bootstrap is version 1.0 operational skill from jwynia/agent-skills that bootstraps agentic development environments from a declarative agent.toml manifest. Running /agent-bootstrap validates the environment and applies tool configuration; /agent-bootstrap --verify runs all verification checks; /agent-bootstrap --check performs a dry run reporting planned changes. Developers reach for agent-bootstrap when spinning up a new coding or task agent repo and need consistent baseline skills, memory structure, and guardrails across Claude Code, OpenCode, and similar agent hosts. The MIT-licensed skill treats infrastructure setup as code so agent projects start with reproducible configuration instead of ad hoc folder layouts.249installs