JoshuaShepherd
- 405 skills
- 0 followers
- 1 repo stars
- 2 weeks ago last updated
- ▌ Asset Animate 2 · joshuashepherdCreate micro-animations from still images — parallax, Ken Burns, floating particles, breathing effects, cinemagraphs. Generates CSS animations, GSAP code, or short video/GIF from a static image. Use for hero backgrounds, social posts, and subtle motion design.
- ▌ Review Scrape 2 · joshuashepherdScrape book reviews from Goodreads and Amazon — fetches actual review text, ratings, dates, and reviewer metadata. Organizes by book, extracts quotable lines, analyzes sentiment patterns, and outputs structured markdown for copy strategy and social proof.
- ▌ Image Optimize 2 · joshuashepherd bundleOptimize images from the local images repo (convert to WebP, generate responsive variants, resize) and upload to Supabase Storage preserving the multi-tenant folder structure. Use when adding new images, syncing the images repo to storage, or generating responsive sets for existing assets.
- ▌ Paratext Audit 2 · joshuashepherdAudit paratext (supporting content) across all platform surfaces — courses, books, articles, exercises, field experiments, and reflection questions. Reports presence, voice quality, rendering gaps, and SEO/GEO readiness. Use before publishing content or to find what needs authoring.
- ▌ Asset Composite 2 · joshuashepherdCombine multiple generated images into a single composite — grids, collages, mood boards, before/after comparisons, or series overviews. Use when reviewing batches, creating mood boards, or building multi-image marketing assets.
- ▌ Paratext Author 2 · joshuashepherdWrite missing or stub paratext for any platform surface — books, courses, podcast episodes — in Alan Hirsch's voice. Queries the database for gaps, writes content, and updates via SQL. Use after running /paratext-audit to fill what it found.
- ▌ Movement Leader Substrate 2 · joshuashepherdAuthor the machine-readable substrate document for a movement leader — identity resolution, relational traversal, conceptual indexing, voice fidelity. Produces a single `{SLUG}_RESEARCH_COLLATED.md` matching the canonical substrate schema. Use when creating or refreshing the primary research artifact for a movement leader. NOT a consulting deliverable — no fit scores, gap analyses, or marketing playbooks.
- ▌ Movemental Welcome Letter 2 · joshuashepherd bundleDraft the Movemental dashboard welcome letter for a specific movement leader from their fragmentation-story dossier. Use whenever the operator says "write the welcome letter for <author>", "draft the dashboard welcome for <leader>", "generate the fragmentation welcome letter", "do a welcome letter for <slug>", or any phrasing that pairs a movement-leader name with the words welcome, letter, or dashboard intro. Reads the leader's research dossier under docs/movement_leader_research/<slug>/ and emits a 1,000–1,050-word, seven-move letter dense with the leader's actual proper nouns — book titles, podcast appearances, partner publications, and the named city of a future practitioner. Save-to-disk is on by default. Pair with movemental-welcome-letter-publish to ship the finished letter to the database.
- ▌ Asset Brand Check 2 · joshuashepherdAudit generated image assets against the platform's brand guidelines. Use before publishing to verify color palette, typography, tone, and visual consistency.
- ▌ Asset Product Shot 2 · joshuashepherdGenerate clean product/object photography prompts using Nano Banana 2. Use for book mockups, course materials, merchandise, or any object that needs a professional studio-quality shot on white or contextual background.
- ▌ Asset Text Overlay 2 · joshuashepherdGenerate images with precise text rendering using Nano Banana 2. Use for social cards, OG images, course certificates, marketing banners, and any asset that needs legible text.
- ▌ Tam Reflected Understanding 2 · joshuashepherdGenerate a reflected understanding document for a movement leader — a second-person mirror of their calling, audience, content, constraints, and credibility
- ▌ Asset Exploded View 2 · joshuashepherdGenerate deconstructed/exploded view prompts using Nano Banana 2. Use for scroll-stop content, framework visualizations, concept illustrations, or any object elegantly taken apart into its components.
- ▌ Asset Prompt Library 2 · joshuashepherdManage reusable NB2 prompt templates for recurring asset types. Use to store, retrieve, list, or customize prompt templates for consistent asset generation.
- ▌ Movemental Committed Voice Bio 2 · joshuashepherdGenerate the Committed Voices onboarding entry for a specific movement leader from their research dossier. Use when asked to "draft the committed-voice bio for <leader>", "generate Committed Voices entry for <slug>", or "extract committed-voices markdown for <author>". Reads the leader's research dossier under movemental-ai/docs/movement_leader_research/<slug>/ and emits one structured markdown file (YAML frontmatter + slightly-warm third-person editorial bio) that maps directly onto the CommittedVoice type rendered at /onboarding/committed-voices.
- ▌ Movemental Leader Corpus Upload 2 · joshuashepherd bundleLoad a movement leader's research into Supabase movement_leader_corpus_data and the related onboarding tables (corpus_review_items, onboarding_tasks, movement_leader_welcome_letters, organizations.onboarding_state). Runs the research:bridge → etl → publish → publish-letter → verify pipeline that compiles the on-disk research tree (profile/, content/, network/, reflected-understanding) into the typed JSONB columns (identity, biography, theology, voice_analysis, calling_profile, books, articles, audio, videos, frameworks, organizations, reflected_understanding_md). Use this WHENEVER asked to push/load/sync/ETL a movement leader's research or dossier into Supabase, wire up movement_leader_corpus_data, fill empty dossier dimensions (videos, audio, frameworks, "A Letter"/reflected understanding, theology, calling), re-run the corpus pipeline, or onboard a leader's corpus — even if the request says "dossier" or "/profile" rather than "corpus_data". Knows the file→column→parser map and the critical substrate rule (s
- ▌ Movement Leader Media Transcribe 2 · joshuashepherdTranscribe YouTube and podcast items from a movement leader's harvest-manifest.json — yt-dlp auto-captions for video, Whisper for podcast RSS/mp3 URLs. Writes verbatim transcripts to docs/harvest/transcripts/ and updates the manifest. Run after movement-leader-content-harvest and before movement-leader-harvest-ingest when LINK/TRANSCRIBE media needs text.
- ▌ Movemental Welcome Letter Publish 2 · joshuashepherdPublish a finished Movemental dashboard welcome letter to the production Supabase database so the named movement leader sees it on their next dashboard load. Use whenever the user says "publish the welcome letter for <author>", "push <author>'s welcome letter live", "ship the dashboard letter", "publish welcome letter from <file>", or any phrasing that combines a finished welcome letter with making it visible to the leader. Companion to the movemental-welcome-letter skill, which generates the draft; this skill takes the finished draft (file or pasted text) and wires it into the right tenant. Run this after the human has read the draft and is ready to make it live.
- ▌ Icon Audit 3 · joshuashepherdAudit and fix icon usage for movemental — mixed libraries, wrong sizes, hardcoded colors, missing accessibility, and illustration anti-patterns. Enforces lucide-react as the single icon library.
- ▌ Color Audit 4 · joshuashepherdExpert color palette audit for movemental — verify semantic token completeness, WCAG contrast compliance, 60-30-10 distribution, and tonal stacking correctness. Light-primary site with regional Midnight sections.
- ▌ Design Audit 3 · joshuashepherdAudit a page or component against the Digital Curator design spec (DESIGN.md) and movemental project conventions. Use to check visual quality, accessibility, token compliance, and consistency before shipping.
- ▌ Responsive Audit 4 · joshuashepherdAudit and fix responsiveness issues for movemental — breakpoint coverage, layout collapse, touch targets, typography scaling, overflow, and navigation. Enforces mobile-first design for Next.js 16 + Tailwind v4.
- ▌ Claude API 3 · joshuashepherdBuild, refine, and master agents using the Anthropic Claude API and @anthropic-ai/sdk. Covers messages API, tool use, extended thinking, prompt caching, streaming, and multi-agent patterns. Fetches live docs from docs.anthropic.com. Use when building or debugging anything on the Claude API.
- ▌ UI Hook Wiring Audit 2 · joshuashepherdAudit and fix UI component data wiring — verifies every component renders the correct data by tracing rendered fields back to the type-safe hooks layer. Checks that book thumbnails, author profiles, course cards, video metadata, and all other domain data reach the correct UI locations through the right simplified or custom hooks. Reports status for each component, then fixes only incorrectly wired paths.
- ▌ Visual Storytelling Audit 2 · joshuashepherdCross-platform audit of visual storytelling components — cards, grids, stats, comparisons, numbered steps, section rhythm, and narrative flow. Ensures every visual element tells the story with the best design possible. Use when auditing pages for narrative clarity, visual hierarchy, and design token compliance.
- ▌ Agent Context 2 · joshuashepherd bundleBuild or modify user context systems — profile assembly, RunContract payloads, context schemas, catalog validation. Use when working on how user data flows into agents.
- ▌ Git Land Main 2 · joshuashepherd bundleSafely land one or many topic branches onto main via GitHub PR: inventory, commit leftovers, push, open/update PR, verify (hosted Actions if it actually ran, else a local gate), merge or squash (repo norm), sync local main, then delete local/remote branches only after main contains the work. Use when the user asks to merge/squash/commit/PR/push to main, land a branch, land multiple branches, clean up merged branches, or close out feature/fix/chore/slice branches. Never commit or push directly to main; never force-push main; never delete a branch until origin/main has it. GitHub Actions billing is not a merge blocker; do not make repos public to get free minutes.
- ▌ Gpt Export 2 · joshuashepherdGenerate 21 markdown knowledge files + custom instructions for creating a Custom GPT strategic consultant from the current workspace project.
- ▌ Page Audit 2 · joshuashepherdHolistic audit of any movemental.com page — UI, content, architecture, UX, and conversion — writes a markdown report and a standalone fix prompt.
- ▌ Summarize 2 · joshuashepherdSummarizes, condenses, or extracts text, transcripts, and key points from URLs, articles, web pages, PDFs, podcasts, YouTube videos, and local files. Handles transcript extraction, transcript-to-document conversion, and acts as a fallback transcription tool when the user asks to transcribe, digest, recap, or get the gist of a link, video, or document.
- ▌ Article Plan 2 · joshuashepherdStrategically plan a single evergreen article — pillar assignment, keyword targeting, section outline, corpus references, internal link targets, and CTA. Run this before article-author when you want to align on strategy before writing. Use when deciding what to write or how to structure a specific article.
- ▌ Chat UI Audit 2 · joshuashepherdAudit and fix chat-based AI UI (chatbots, agents, floating chat) against production best practices from Claude, ChatGPT, and modern conversational AI interfaces. Use when building, reviewing, or improving any chat UI.
- ▌ Studio Export 2 · joshuashepherdGenerate a feature constitution — data model, product story, and acceptance criteria for a platform slice/feature. Designed for pasting into AI Studio Build mode or any prototyping tool. No styling or UI constraints.
- ▌
- ▌ Movemental PDF 2 · joshuashepherd bundleProduce branded Movemental PDFs (the Movemental, LLC visual identity used at movemental.ai — wordmark logo, Inter typography, monochrome palette with warm-gray accents, generous whitespace, brand panels and stat cards). Use this skill whenever the user asks for a Movemental-branded document, a Movemental Field Guide, a Movemental MOU or proposal, a Movemental report, whitepaper, or any PDF matching the movemental.ai aesthetic. Use it even when the user only says "make this look like our brand", "the way we did it last time", or "in our house style" if the context is Movemental, and any time the deliverable is a PDF for Joshua Shepherd or Movemental, LLC. The skill includes the logo, Inter web fonts, a component library (cover, section openers, brand panels, stat cards, comparison tables, layer pages, checklists, framework callouts, citations, colophon), and a two-pass Chromium render pipeline that produces print-ready PDFs with running header, page-numbered footer, and a suppressed-chrome cover page.
- ▌ Article Audit 2 · joshuashepherdAudit a complete article against all criteria — Alan's five voice markers, SEO/GEO requirements, section architecture, corpus citation quality, pillar alignment, and CTA. Returns per-criterion scores, specific failing passages, and a prioritized revision list. Use before publishing any article, or to diagnose why an article isn't performing.
- ▌ Docs Type Safety 2 · joshuashepherdAudit and update the type safety documentation in _docs/type/ to match the actual codebase state. Ensures docs are the single source of truth without errors or gaps.
- ▌ Audio Scrape 2 · joshuashepherdScrape podcast episodes, sermons, and audio sources — discovers via iTunes Search API, parses RSS feeds, downloads audio, transcribes via OpenAI Whisper, chunks for search, and upserts into the podcast_episodes table.
- ▌ Article Author 3 · joshuashepherdWrite a complete evergreen pillar article in Alan Hirsch's voice — with corpus research from his books, full SEO/GEO architecture, and proper citations. This is the main writing skill. It runs planning, corpus research, and writing in sequence. Use when writing any new article for the platform.
- ▌ Network Map 2 · joshuashepherdMap relationships, co-authorships, endorsements, and organizational connections across movement leaders. Reads existing per-leader research in docs/movement_leader_research/ and writes a cross-leader graph. Output table schemas align with the `movement-leader-substrate` Network graph section so per-leader substrate files can be regenerated from this data.
- ▌ Audit Experience 2 · joshuashepherdAudit the course learn experience for section coverage, sidebar rendering, layout, design token compliance, tenant isolation, accessibility, data flow, and component quality. Use when reviewing course UI code quality or diagnosing learn experience issues. Pass a course slug or "all".
- ▌ Validate 2 · joshuashepherdRun bottom-up layer validation and fix all errors. Use before starting work on any layer or to check project health. Follow the Lock-Before-Proceed and Error Fixing Protocol.
- ▌ Gemini API 2 · joshuashepherdBuild, refine, and master agents using the Google Gemini API and @google/generative-ai SDK. Covers generateContent, function calling, Google Search grounding, multimodal input (text/image/audio/video), streaming, context caching, and the Gemini 2.5 and 3.x model families. Fetches live docs from ai.google.dev. Use when building or debugging anything on Google Gemini.
- ▌ Openai API 2 · joshuashepherdBuild, refine, and master agents using the OpenAI API — Chat Completions, Responses API, and OpenAI Agents SDK. Covers function calling, structured output, streaming, multi-agent handoffs, and the full gpt-5.4 family. Fetches live docs from platform.openai.com and openai.github.io/openai-agents-python. Use when building or debugging anything on OpenAI.
- ▌ Youtube Scrape 2 · joshuashepherdScrape YouTube channel videos — fetches metadata, downloads transcripts via yt-dlp, chunks for search, and pulls comments. Upserts into the videos table.
- ▌ Env Setup 2 · joshuashepherdScaffold and validate environment variables for Next.js 15 or Vite + Express — Zod-validated env schema, .env.example with all required vars, runtime validation that fails fast on missing values, and a check:env script. Use when starting a new project or auditing env var hygiene.
- ▌ Trigger Dev 2 · joshuashepherd bundleWork with Trigger.dev (v4) — the open-source background jobs framework for writing durable tasks, scheduled/cron jobs, waits, Realtime, React hooks, and chat.agent AI agents in TypeScript. Use when the user mentions Trigger.dev, writes or triggers tasks, sets up trigger.config.ts, uses @trigger.dev/sdk, the trigger.dev CLI, the Management API, or asks how background jobs / queues / retries work. ALWAYS fetch current docs before writing code.
- ▌ Design Chain Audit 2 · joshuashepherdUnified design chain audit for Movemental — flushes drift across all 5 layers (Stitch → tokens → primitives → components → layouts) covering color, typography, Tailwind hygiene, pattern, a11y, and interaction fidelity against the "Warm Scholarly Authority" design system. Use before shipping any UI or after a large build sprint.
- ▌ Voice Designer 2 · joshuashepherdBuild detailed voice profiles for characters, narrators, brands, or AI agents — covering diction, syntax, rhythm, worldview, signature patterns, and failure modes. Use when creating a character voice, designing an agent's persona, or auditing voice consistency.
- ▌ Youtube Transcript 2 · joshuashepherdExtract a plain-text transcript from any YouTube video URL. Returns timestamped and full-text versions of the transcript saved to a local file.
- ▌ Affiliation Audit 2 · joshuashepherdAudit collected affiliation data against logo strip and social proof best practices. Evaluates logo quality, grouping strategy, copy framing, prominence ordering, and strip-readiness. Run after affiliation-scrape, before logo-strip-author.
- ▌ Agent RAG 2 · joshuashepherd bundleBuild or modify a RAG retrieval pipeline — vector store search, corpus routing, citation rendering, book fidelity. Use when adding retrieval to agents.
- ▌ Affiliation Scrape 3 · joshuashepherdResearch organizations a leader is affiliated with AND fetch logos for them. Output is a JSON record optimized for rendering a logo strip / social-proof section on the platform site. NOT the same as substrate organizational affiliations — for the substrate's structured org list, use `movement-leader-substrate`.
- ▌ Agent Test 2 · joshuashepherd bundleWrite tests for agents — unit tests for tools/instructions/routing, smoke tests for end-to-end behavior, e2e specs for API contracts. Use when adding test coverage to agents.
- ▌ Agent Tool 2 · joshuashepherd bundleCreate or modify an OpenAI Agents SDK tool with Zod params, caching, metrics, and agent registration. Use when adding tools to agents.
- ▌ Asset Edit 2 · joshuashepherdEdit or refine an existing image using Nano Banana 2's conversational editing. Use when an asset needs color correction, element changes, style adjustments, or iterative refinement.
- ▌ Fragmentation Story 2 · joshuashepherdTell a named movement leader's actual story of informational and relational fragmentation. Inventories every place their content currently lives — books, PDFs, articles, blogs, websites, social media, their own head, notes/laptop, apps, podcasts, YouTube, sermons, talks — with specific titles and locations pulled from prior research. Use when asked to map fragmentation for a leader, audit where their content is scattered, or produce a "where does it all live" report. Reads from `docs/movement_leader_research/{slug}/` and writes `fragmentation-story.md` into that leader's folder.
- ▌ Tam Headshot Source 2 · joshuashepherdSurgically acquire one verified headshot per movement leader — official sources first, identity checks, direct download, manifest line — without bulk scraping or image-search roulette. Use when building TAM avatars, profile media, or filling knowledge leader folders.
- ▌ Asset Match 2 · joshuashepherdAudit where images are needed across the project (heroes, covers, cards, OG images) vs what's available in public/images/. Reports matches, mismatches, missing assets, and broken references. Connects to asset generation skills for creating missing images.
- ▌ Tailwind Cleanup General · joshuashepherdRun a best-practice Tailwind cleanup that stays **inside the design system this repo already uses**. This skill is intentionally **design-agnostic**: it does not assume shadcn, Material-3, an "ink" system, CSS-variable arbitrary values, or any particular palette. It discovers what the repo does, treats the *best* of that as the standard, documents it if it isn't already, and then makes the code consistent with it.
- ▌ Agent Stream 2 · joshuashepherd bundleWork on SSE streaming, plain-text transport, ChatKit events, and session continuity. Use when debugging or modifying agent streaming behavior.
- ▌ Asset Mockup 2 · joshuashepherdPlace generated images into real-world context mockups — laptop screens, phone frames, book mockups, poster frames, billboard mockups. Use for marketing previews, social media content, and stakeholder presentations.
- ▌ Asset Series 2 · joshuashepherdGenerate a visually consistent set of image assets using Nano Banana 2. Use for course module covers, book chapter headers, video series thumbnails, or any batch that needs visual continuity.
- ▌ Figma Prompt 2 · joshuashepherdGenerate a complete Figma Make prompt package — Context Primer, Checkpoint, Section Build Prompts, and Iteration Templates — for any UI page, component, or screen. Uses intent-first prompting and Figma-native vocabulary. Counterpart to the studio-prompt skill, tuned for design artifacts rather than working code.
- ▌ Stitch React 2 · joshuashepherdConvert Stitch screens to React components — extracts design tokens, decomposes HTML into shadcn/ui components, generates TypeScript types, wires Layer 5 hooks, and archives existing pages. Triggers on: Stitch React, component conversion, React conversion, HTML to React. NOT for: new React apps, API routes, services, or schema changes.
- ▌ Type Safety Chain 2 · joshuashepherdImplement the full six-layer type safety chain in a new project — DB → Drizzle schema → Zod schemas → services → API routes → React hooks. Installs dependencies, writes all infrastructure files, generates all layers from the live database, and validates each layer bottom-up. Use for new projects that need this architecture from scratch, or to verify and repair an existing chain. Supabase MCP required for DB verification.
- ▌ Coding Standards 2 · joshuashepherd bundleUniversal coding standards, best practices, and patterns for TypeScript, JavaScript, React, and Node.js development.
- ▌ Alan Voice 2 · joshuashepherdWrite, edit, or audit content in Alan Hirsch's exact voice — using the same five voice markers, argument patterns, rhetorical posture, and failure mode guards as the AI Lab agent. Use when writing any published content, course material, marketing copy, agent prompts, or social posts that must sound authentically like Alan.
- ▌ Agent Create 2 · joshuashepherd bundleScaffold a new OpenAI Agents SDK agent with instructions, tools, API route, and bridge wiring. Use when creating a new specialized agent from scratch.
- ▌ Home Consult 2 · joshuashepherdStrategic consultation for a movement leader home page — mandatory sections rubric, scroll-stop strategy (Spline/NB2/GSAP), copy evaluation, and conversion architecture. Designed for thought leaders in the missional/formation niche.
- ▌ Course Author 2 · joshuashepherdGenerate course content for a specific section of the transformation loop, following the actual mDNA course structure, Charter requirements, and Alan Hirsch's voice. Use when writing individual course sections.
- ▌ Course Ingest 2 · joshuashepherdIngest course content from markdown files into the database. Parses manifest + module files, validates, and upserts course_lessons, reflection_questions, discussion_prompts, and exercises.
- ▌ Pathway Audit 2 · joshuashepherdAudit a pathway page against the canonical 12-section architecture. Reports what's missing, what's below spec, what's complete, and what needs improvement for SEO/GEO, voice, and formation quality. Use before publishing a pathway or after adding new sections.
- ▌ Article Author 2 · joshuashepherdWrite a complete evergreen pillar article in Alan Hirsch's voice — with corpus research from his books, full SEO/GEO architecture, and proper citations. This is the main writing skill. It runs planning, corpus research, and writing in sequence. Use when writing any new article for the platform.
- ▌ Article Corpus 2 · joshuashepherdLook up passages, themes, and arguments from Alan Hirsch's book corpus — both from the local markdown files and the Supabase database. Use when researching what Alan actually says about a topic before writing an article. Returns quoted passages with proper citations.
- ▌ Asset Generate 2 · joshuashepherdGenerate image assets using Nano Banana 2 (Gemini Flash Image). Use when creating hero images, course covers, book covers, thumbnails, social cards, or any visual asset for the platform.
- ▌ Pathway Author 2 · joshuashepherdWrite or update content for a pathway page — any single section or a full pathway. Follows the canonical 12-section architecture, Alan Hirsch's voice, and SEO/GEO best practices. Use when writing new pathways, filling missing sections, or improving existing content.
- ▌ Course Scaffold 2 · joshuashepherdScaffold a new 8-week course with all canonical section types, database rows, and optional markdown file structure. Use when creating a new course from scratch.
- ▌ Course Validate 2 · joshuashepherdValidate a course against the Transformational Course Charter — checks Four Necessities, section completeness, word counts, and structure. Use before publishing or after authoring.
- ▌ Logo Strip Author 2 · joshuashepherdAuthor a complete social proof / logo strip section for an author or leader — copy, grouping strategy, TypeScript content definition, React component spec, CSS/Tailwind implementation, and logo download checklist. Run after affiliation-scrape and affiliation-audit.
- ▌ Affiliation Scrape 2 · joshuashepherdResearch and document all organizations a leader/author is affiliated with — publishers, speaking bureaus, conferences, denominations, media outlets, endorsing ministries, and academic institutions. Uses browser search to build a structured affiliation record for use in logo strips and social proof sections.
- ▌ Agent Instructions 2 · joshuashepherd bundleAuthor or refine agent system prompts using composable instruction layers (identity, theme, mode, style, context). Use when writing or tuning agent behavior.
- ▌ Remotion 2 · joshuashepherd bundleProgrammatic video creation with React via Remotion — compositions, animations, motion graphics, rendered to MP4. USE WHEN video, animation, motion graphics, video rendering, React video, intro video, YouTube video, TikTok video, video production, render video, content to animation, animate content, video overlay.
- ▌ Video Consult 2 · joshuashepherdConsult on video production for a course — scripture grounding, narrative arc, storyboard, scene structure, and script. Bridges Alan's voice, course theology, and Remotion video production. Use when planning or authoring video content for any course.
- ▌ Video Researcher 2 · joshuashepherdResearch real-world statistics, insights, and compelling data about online courses, completion rates, learning effectiveness, and transformative education — then organize findings into suggested video components, script lines, and data visualizations for course explainer videos. Use before scripting or updating video content.
- ▌ Icon Audit 2 · joshuashepherdAudit and fix icon and illustration usage across the codebase — mixed libraries, wrong sizes, hardcoded colors, missing accessibility, flex compression, and illustration anti-patterns. Run before any design review.
- ▌ Color Audit 2 · joshuashepherdExpert color palette audit — verify light/dark mode alignment, WCAG contrast compliance, 60-30-10 distribution, and overall color harmony. Checks the palette itself, not just usage.
- ▌ Design Audit 2 · joshuashepherdAudit a page or component against the AntiGravity 5-dimension design framework and project conventions. Use to check visual quality, accessibility, and consistency before shipping.
- ▌ Tenant Check 2 · joshuashepherdAudit components for hardcoded tenant strings, non-semantic colors, or missing feature flags. Use before PR review or to check tenant isolation.
- ▌ Movement Leader Voice 2 · joshuashepherdAuthor the canonical voice & style guide for a movement leader at `docs/voice/{SLUG}_VOICE.md` — a self-contained, corpus-grounded reference (five weighted voice markers, prose character, signature rhetoric, failure modes, argument patterns, signature phrases, citation habits, audit rubric, and sample passages) describing how this leader sounds, on their own terms, not by comparison to anyone else. Run inside a leader's website repo against `docs/books/`, `docs/movement_leader_research/`, and any other content. Use when standing up or refreshing a leader's voice identity for Writing Studio, AI Lab, agents, and human editors. Distinct from the Movemental company voice ([[movemental-voice]]) and the generic [[voice-designer]]. Reference implementation is the brad-brisco repo.
- ▌ Movement Leader Themes 2 · joshuashepherdDerive a movement leader's core themes from their own book corpus and research, then author the canonical theme docs in `docs/themes/` — a `CORE_THEMES.md` taxonomy (which themes earn pathway status, and why) plus one deep-dive `{slug}.md` per theme (12-section pathway content, grounded and cited). Run inside a leader's website repo against `docs/books/`, `docs/movement_leader_research/`, and any other content. Use when standing up or refreshing the theme system for a movement leader. Self-contained; reference implementation is the brad-brisco repo.
- ▌ Design Section 2 · joshuashepherdDesign a new UI section or component using the 5-dimension AntiGravity framework (pattern, style, color, typography, animation). Use when building a new visual section, hero, feature grid, or page layout.
- ▌ Responsive Audit 2 · joshuashepherdAudit and fix responsiveness issues across any section or page — breakpoint coverage, layout collapse, touch targets, typography scaling, overflow, image sizing, and navigation behavior. Enforces mobile-first design for this Next.js + Tailwind + shadcn platform.
- ▌ Tailwind Cleanup 2 · joshuashepherdScan and fix Tailwind anti-patterns — hardcoded colors, arbitrary values, raw HTML bypassing shadcn, layer violations, and dark/light mode issues. Enforces the D1→D6 token-first design chain.
- ▌ Typography Polish 2 · joshuashepherdAudit and fix prose/content typography across all rendering surfaces — articles, books, courses, chat responses, and any component rendering HTML or markdown from the database. Ensures beautiful, consistent long-form reading experiences.
- ▌ Claude API 2 · joshuashepherd bundleBuild apps with the Claude API or Anthropic SDK. TRIGGER when: code imports `anthropic`/`@anthropic-ai/sdk`/`claude_agent_sdk`, or user asks to use Claude API, Anthropic SDKs, or Agent SDK. DO NOT TRIGGER when: code imports `openai`/other AI SDK, general programming, or ML/data-science tasks.
- ▌ Movement Leader Platform Build 2 · joshuashepherd bundleAudits a movement-leader tenant against the master component checklist, writes a gap-remediation build prompt to docs/build/prompts/, then implements missing pages and components in the repo's existing design system with full type-safety compliance. Use when running the platform checklist, auditing component gaps, building remaining pages, platform build, master-component-checklist, or remediating missing shell/themes/ library/formation routes. Run from any movement-leader repo root.
- ▌ Color Audit 3 · joshuashepherdAudit color palette for light/dark mode alignment, WCAG contrast compliance, and consistent token usage.
- ▌ Responsive Audit 3 · joshuashepherdAudit pages for responsiveness across mobile, tablet, and desktop breakpoints. Check layout, typography, overflow, touch targets, and spacing.
- ▌ Tailwind Cleanup 3 · joshuashepherdScan and fix Tailwind anti-patterns — hardcoded colors, arbitrary values, inconsistent spacing, and dark/light mode issues.