Every skill across the ecosystem, generated 2026-07-26. Tap a row to expand. Search and filters combine.
115 shown
TJ Meaney job application log and workflow
kindly-agents/agents/gretta/skills/job-search-detailed/applications.md
TJ Meaney's profile, target roles, key stats, and differentiators for job search
kindly-agents/agents/gretta/skills/job-search-detailed/candidate.md
Update all relevant files after any client interaction
kindly-agents/agents/gretta/skills/client-update.md
Formats every kindly-comms message to TJ for scannability, headline-first, front-loaded, 1-3 short sentences, depth pushed to accordions or cards. Load BEFORE composing any kindly-comms reply, accordion, card, or brief longer than a one-line ack, and when deciding reply vs accordion vs card. Grounded in NN/g reading research (people scan, they do not read) plus TJ's 2026-07-01 correction. Do NOT use for Telegram (telegram-communication skill) or for drafts in TJ's own voice (CONTEXT.md TJ-voice rules).
kindly-agents/agents/gretta/skills/comms-formatting.md
Choose the right diagram format and generate it on request or when a visual would help
kindly-agents/agents/gretta/skills/diagram-on-demand.md
TJ Meaney job search skill — entry point and task router
kindly-agents/agents/gretta/skills/job-search-detailed/index.md
Tailoring resumes and cover letters for TJ's job applications, scraping company job boards, and building application materials.
kindly-agents/agents/gretta/skills/job-search.md
Extracted from SOUL.md 2026-07-26 (context-rightsizing pass, TJ-approved). SOUL keeps the four-line intent; the how-to lives here. Load when actually running one of these procedures.
kindly-agents/agents/gretta/skills/learning-loop.md
Onboarding guide for setting up a personal AI assistant using Claude Desktop (Cowork), Obsidian, and a structured knowledge base. No code required.
kindly-agents/agents/gretta/skills/level-1-cowork-setup.md
TJ Meaney LinkedIn profile — full copy-paste ready, optimized April 2026
kindly-agents/agents/gretta/skills/job-search-detailed/linkedin.md
- TJ asks for a wealth/career/venture architecture check ("run the Naval audit", "am I building wealth or just busy").
kindly-agents/agents/gretta/skills/naval-wealth-audit.md
TJ Meaney canonical resume content — source of truth for all tailored versions
kindly-agents/agents/gretta/skills/job-search-detailed/resume.md
Lead TJ through the Sunday weekly planning session at 2pm MDT argument-hint: "[optional: specific focus or context]"
kindly-agents/agents/gretta/skills/sunday-planning.md
Capture a task or open loop from conversation into the right system
kindly-agents/agents/gretta/skills/task-capture.md
Cover letter structure, anti-AI writing rules, ATS checklist, and job boards for TJ Meaney
kindly-agents/agents/gretta/skills/job-search-detailed/writing-rules.md
Orchestrates the creation of a new Kindly agent — interview to gather identity + scope + type (worker/daemon vs. resident), call the kindly-agents scaffolder, kick off the first-boot ONBOARDING.md interview. Use when TJ asks for a new agent.
kindly-tools/skills/agent-ops/agent-creation-helper/SKILL.md
Judge when to act autonomously versus stop and ask the user. Use when deciding whether an action needs approval or calibrating how much independence to take on a task. (Doctrine - candidate for relocation into agent CLAUDE.md/SOUL docs; see TODO.)
kindly-tools/skills/agent-ops/autonomy-control/SKILL.md
Runs a feature build as an orchestrated multi-agent pipeline (the explore, plan, parallel-implement, merge, review, PR flow): parallel Sonnet explorers map the repo, the top model writes one plan, implementers execute plan slices in isolated git worktrees, a merge stage combines them, a high-effort reviewer adversarially checks the diff with a fix loop, and the output is a branch plus draft PR for TJ. Use when the user says "run the build pipeline", "build this with the pipeline", "use the build flow", "multi-agent build", "swarm this feature", or hands over a feature/change big enough to split across parallel implementers (3+ files or 2+ independent slices). Do NOT use for one-file edits, quick fixes, or pure research (use the Workflow tool ad hoc), and never for repos where the CLAUDE.md forbids agent builds.
kindly-tools/skills/agent-ops/build-pipeline/SKILL.md
Process for creating a client design system repo from brand assets gathered at onboarding — colors, fonts, logos, CSS tokens, and agent-invocable SKILL.md
kindly-tools/skills/agent-ops/client-design-onboarding/SKILL.md
Audits and rewrites repo context (CLAUDE.md files, SKILL.md files, agent boot docs, system prompts) to the Claude-5-generation context-engineering rules from Anthropic's 2026-07 article, trust model judgment over micro-rules, progressive disclosure over upfront walls, interface design over examples, gotchas over the obvious. Use when the user says "update repo context", "rightsize CLAUDE.md", "context engineering pass", "run the context skill", "unhobble", "apply the Claude 5 context rules", "doctor my context", or when authoring/overhauling context for agents on Opus 5 / Fable 5 generation models. Produces per-file diff proposals behind an approval gate, never silent rewrites. Do NOT use for general knowledge-doc pruning with no model-generation angle (writing-knowledge-docs) or for authoring a brand-new skill (creating-skills).
kindly-tools/skills/agent-ops/context-rightsizing/SKILL.md
Convene a multi-agent review council that pressure-tests a decision, plan, idea, build, or architecture, then synthesizes one verdict. Spawns several sub-agents with different lenses (thinking-styles for ideas/plans, retuned thinking-styles for architecture/technical decisions, domain experts for web builds), has them peer-review each other anonymously, and a chairman produces a final recommendation. Use when the user says "council this", "run the council", "pressure-test this", "stress-test this", "debate this", "war room this", "design-review this", "council this architecture", "review this build", "review this plan", "review this flow", "should I X or Y", "which is stronger", "I'm torn between", "what would you do". Built on Andrej Karpathy's LLM Council methodology. NOT for one-right-answer questions, creation tasks, or trivial choices.
kindly-tools/skills/agent-ops/council/SKILL.md
Authors new agent skills (SKILL.md files) through a short interview, then writes, tests, and registers them. Use whenever the user says "make a skill", "create a skill", "turn this into a skill", "skill-ify this", "add a skill for X", or asks how to write, fix, improve, or debug a SKILL.md, including skills that never trigger. Do NOT use for scheduled/cron pipelines (use creating-workflows) or for one-off instructions that belong in CLAUDE.md.
kindly-tools/skills/agent-ops/creating-skills/SKILL.md
Build and understand agent workflows end to end. What a workflow is, how to author one (GOAL.md + WORKFLOW.md), the file formats, how workflows execute, and where the model is going (Homestead-owned workflows, spec 0020). Use whenever the user asks to build, create, author, add, schedule, or set up a new workflow, cron job, recurring task, or scheduled pipeline for an agent, OR asks how a workflow runs, why a cron did not fire, or how the execution model works. Absorbed the former workflow-standard reference 2026-06-18.
kindly-tools/skills/agent-ops/creating-workflows/SKILL.md
13-step structured reasoning framework for attacking complex problems — covers problem definition, inversion, root cause analysis, MECE, second-order thinking, and validation
kindly-tools/skills/agent-ops/critical-thinking/SKILL.md
Diagnose and resolve drift between TJ's M4 dev computer and the Meaneyserver. Walks every git repo under ~/kindly-projects/, reports ahead/behind/uncommitted state per repo, identifies which side has the canonical state, proposes surgical actions per repo (pull/push/stash/merge/rebase), gets TJ's approval per destructive action, never blanket-commits or force-pushes without explicit confirmation.
kindly-tools/skills/agent-ops/cross-machine-sync/SKILL.md
Sanity-check across the entire Kindly Creative codebase — verify all client repos conform to the canonical structure, all agents follow the agent-template shape, no agency docs duplicated into projects, no per-project cron registries, no hardcoded credentials in tracked files. Surface drift, propose surgical fixes, never blanket-commit.
kindly-tools/skills/agent-ops/ecosystem-audit/SKILL.md
Standard pattern for reporting unrecoverable errors from agents to TJ. Use any time a workflow, skill, or requested task hits an error that the agent cannot resolve itself. Covers: what to send TJ (plain human message, not [CRON:] gibberish), what to log (to run.log), when to escalate via notify-ops.py, and when to stay silent (transient/retriable failures). Use whenever the user's last message triggered work that went wrong, OR when a scheduled workflow hits an unrecoverable issue.
kindly-tools/skills/agent-ops/failure-reporting/SKILL.md
Convene the Kindly Creative Council, TJ's standing panel of business-specialist advisors (pricing, client portfolio, positioning, craft, AI-ops, growth) that score and pressure-test a real business decision, loop until the verdict stabilizes, and log the outcome. Use when TJ says "kindly council this", "run this by the council", "run the business council", "convene the council", "ask my council", "score this with the council", or wants the council's take on a pricing call, taking or dropping a client, a new service line or offer, positioning, or a product bet like the health app. For real cross-functional BUSINESS decisions only. Do NOT use for a fast multi-lens gut-check (use quick-council), ad-hoc thinking-style review of a non-business decision, plan, or architecture (use council), or testing creative against customer personas (use persona-panel).
kindly-tools/skills/agent-ops/kindly-council/SKILL.md
Decide what an agent should write to memory, what to let go, and how to keep memory useful instead of noisy. Use when managing agent memory, deciding what to persist, or pruning stale notes. (Doctrine - candidate for relocation into agent CLAUDE.md/SOUL docs; see TODO.)
kindly-tools/skills/agent-ops/memory-management/SKILL.md
AI-led interview + scaffold for new Kindly Creative client repos. Conducts a short conversational interview (display name, slug, project type, primary contact, optional kickoff doc and brand-kit location), copies the canonical template from kindly-tools/templates/client-template/, substitutes placeholders, adds conditional folders by project type, extracts content from the kickoff doc to fill ICP / brand voice / value props / offerings where possible, initializes git, registers the client in kindly-tools/KINDLY.md, prints next steps. Use when the user says "onboard a new client", "scaffold X", "set up [client] in kindly-clients/", or starts a new client kickoff.
kindly-tools/skills/agent-ops/onboarding-client/SKILL.md
Run any question or decision through five advisors with different thinking styles, then anonymized peer review, then a Chairman verdict, all in ONE pass in the current context (no sub-agents, no report file). The fast, everyday version of the council, for smaller or quicker decisions where you want multiple angles without the heavyweight machinery. TRIGGERS, fire on any of these: "quick council", "fast council", "run a quick council", "lite council", "five advisors", "run the five advisors on this", "council this quickly", "council this real quick", "give me the five lenses", "pressure-test this quickly". For BIG, expensive, or high-stakes decisions that warrant parallel sub-agents + a saved report, use council instead. Do NOT fire on simple factual lookups or pure creation tasks.
kindly-tools/skills/agent-ops/quick-council/SKILL.md
Standard end-of-session ritual every agent runs when wrapping up a working session. Use when the user says "wrap up", "close out", "session end", "log the session", "end of day", "that's it for now", or when a session-close cron trigger fires (e.g. Tue/Wed/Thu 8pm for Gretta). Produces a session log entry, updates the task ledger with status changes, finalizes the day's daily note, flags stale items, and considers curated memory updates. Do NOT confuse with wind-down (the 6:30pm family-time reminder) — session-close happens when the active work session actually ends.
kindly-tools/skills/agent-ops/session-close/SKILL.md
Voice, format, and structural rules for every Telegram message an agent sends to TJ. Use whenever you are about to call the Telegram MCP to send a message, reply, or post an update — including briefs, reminders, flagged items, reports, drafts, and questions. Covers length, formatting (plain text, no markdown headers), banned words, when to break into multiple messages, mobile-read budget, what to skip, and how to end a message. Load BEFORE calling any Telegram send tool if the message is more than a one-line "ok" acknowledgment.
kindly-tools/skills/agent-ops/telegram-communication/SKILL.md
Check work before calling it done - the discipline of verifying your own output (tests, fact-checks, diffs, re-reads) every time. Use when verifying or double-checking completed work. (Doctrine - candidate for relocation into agent CLAUDE.md/SOUL docs; see TODO.)
kindly-tools/skills/agent-ops/verification/SKILL.md
Write and prune agent knowledge docs so they stay lean, high-signal, and behavior-changing. Use when auditing or cleaning up a CLAUDE.md, SOUL, CONTEXT, MEMORY, USER, or any always-loaded knowledge file; when a knowledge doc feels bloated, stale, or duplicated; when writing a new CLAUDE.md from scratch; or when the user says "clean up CLAUDE.md", "audit my knowledge docs", "this file is bloated", "make my memory files lean", "strip down CONTEXT.md", "prune my agent docs", or "is this knowledge doc any good". For authoring a NEW reusable skill (not editing a knowledge doc), use creating-skills instead.
kindly-tools/skills/agent-ops/writing-knowledge-docs/SKILL.md
Brand strategy reference for Kindly Creative LLC. Load when working on Kindly Creative website copy, homepage messaging, service pages, product offering, positioning, brand voice, or any Kindly Creative marketing collateral. Use when TJ says "work on the Kindly Creative website", "update the homepage", "update the messaging", "write copy for Kindly", "work on the product offering", "brand strategy", "Kindly positioning", or "StoryBrand".
kindly-tools/skills/branding/kindly-brand/SKILL.md
Use this skill to generate well-branded interfaces and assets for Kindly Creative LLC, either for production or throwaway prototypes/mocks/decks/slides/UI. Contains essential design guidelines, colors (emerald green accent, warm off-white neutrals), type (Inter site / Playfair Display carousel), fonts, logos, and UI kit components for prototyping the marketing site and social carousels. user-invocable: true
kindly-tools/skills/branding/kindly-design/SKILL.md
Donald Miller's StoryBrand SB7 framework for clarifying brand messaging. Use when writing or auditing homepage copy, landing pages, sales pages, founder bios, value propositions, or any place where a brand needs to position itself as the guide and the customer as the hero. Distinct from landing-page-psychology (which covers temperature/scroll mechanics) and copywriting (which covers voice and headline mechanics) — StoryBrand is the higher-level narrative architecture: who is the hero, what problem do they have, who is the guide, what's the plan, what's the call, what happens if they don't act, what happens if they do.
kindly-tools/skills/branding/storybrand/SKILL.md
Writes AS TJ Meaney, in his own register, grounded in mined primary sources (his kindly-comms chat history, git commit history, and documented voice rules) across four distinct registers, casual SMS to family and friends, work-chat to agents and team, founder-public voice for website copy and founder videos and LinkedIn, and his work-register for reasoning through a decision. Use whenever writing as TJ, for texts he sends, founder copy, website copy in his point of view, LinkedIn posts, video scripts, or outbound email from his address, or any ask phrased as "write this as me," "in my voice," or "from my point of view." Living skill, re-mine and refresh as his voice evolves, never treat as a frozen snapshot. Do NOT use for Gretta's own voice replying to TJ (see telegram-communication) or for Kindly's general brand voice (see kindly-brand), those are different speakers.
kindly-tools/skills/branding/tj-voice/SKILL.md
"AEO content structure: writing and structuring content to win Google AI Overviews, featured snippets, and answer-engine citations. FAQ/HowTo/QAPage schema, extractable answers, entity SEO. Use when structuring content for AI answer visibility (the fix-side companion to aeo-rendering)."
kindly-tools/skills/marketing/aeo-content-structure/SKILL.md
How to make a website discoverable by AI crawlers (ChatGPT/GPTBot, Perplexity, Claude web search, Bing Copilot) when the site is a client-rendered SPA. Covers diagnosis (curl test + view-source check), why client-only React/Vue/Svelte ships empty HTML to non-Google bots, the static fallbacks that DO ship (JSON-LD schema, llms.txt, robots, sitemap), the three fix paths (build-tool prerender plugin / migrate to SSG meta-framework / server-side rendering), the tradeoff matrix between them, quick wins available without pre-rendering, and the default Kindly recommendation. Load before any client conversation about AEO performance, before scoping a website rebuild for AEO, or when seo-aeo-audit flags an empty-HTML problem.
kindly-tools/skills/marketing/aeo-rendering/SKILL.md
"Build a client's AI-search + SEO content strategy blueprint BEFORE any pages get written: a keyword universe ranked by buying intent, competitor keyword gaps, competitor positioning + citation sources, quote-worthy citation-magnet questions, an ideal-reader persona, and a topical authority map with publish order. Every step is grounded in real DataForSEO data, not LLM guessing. Use when the user says 'keyword research', 'keyword strategy', 'topical authority map', 'content plan to rank', 'what should we write', 'AEO/GEO strategy', 'what should we write to get cited by ChatGPT', 'find keyword gaps', 'who are our customers / build personas from search data', or is scoping a content plan for a client (or Kindly itself). This is the THINKING/research front-end producing a full blueprint (not ad-hoc keyword lists, the seo skill covers quick keyword questions); geo and aeo-content-structure are the EXECUTION back-end, use them when the plan exists and the work is optimizing presence or pages."
kindly-tools/skills/marketing/ai-search-strategy/SKILL.md
End-to-end playbook for writing SEO/AEO/GEO blog posts that rank, get cited by AI, and convert. Covers keyword gap pull, topic clusters, intent matching, answer-first structure, conversion mechanics, internal linking, refresh cadence, schema, publishing flow, and performance tracking
kindly-tools/skills/marketing/blog-writing/SKILL.md
'Restructures any draft, a long page, an email sequence, a video script, or a social post, for engagement using the Crumb Trail: hook, gap, reward, deepen, ask. Fixes pacing so people keep going instead of leaving: chunks the draft into zones, opens one curiosity gap at a time, caps CTAs at one per zone, and ends every zone on a named thread instead of a fade-out. Use when TJ says "breadcrumb this," "people will bounce," "make this flow," "funnel logic," "too much at once," or "restructure for engagement." Auto-fire when building any page, email sequence, video script, or long post, even when pacing is not asked for directly. Do NOT use for page section order or the temperature model (landing-page-psychology), CRO testing and benchmarks (conversion-optimization), or line-level headline and CTA wording (copywriting). This skill owns sequencing and load pacing across formats.'
kindly-tools/skills/marketing/breadcrumb-engagement/SKILL.md
CRO methodology, benchmarks, testing, and high-converting page structure. Use when optimizing conversion rate, auditing a page for friction, prioritizing what to test/fix, or structuring a homepage/landing page for lead gen. For the persuasion narrative, temperature model, and section-by-section copy playbook, use landing-page-psychology instead.
kindly-tools/skills/marketing/conversion-optimization/SKILL.md
The shared copywriting foundation - voice rules and kill list, Schwartz awareness levels, frameworks (PAS/AIDA/BAB/FAB), headline and first-person CTA formulas, microcopy, objection-handling, and proof copy, grounded in Ogilvy/Schwartz/Cialdini with 2026 data. Load as the mechanics layer for any copy task. To pick the angle, pair it with copywriting-sales (direct offer, desired action) or copywriting-awareness (brand-building or education, no offer).
kindly-tools/skills/marketing/copywriting/SKILL.md
Write brand-building, audience-educating copy for organic social, Threads, Instagram carousels, and thought-leadership pieces - the awareness angle, hook, and narrative voice when there is no direct offer. Use for awareness/top-of-funnel content, "write something about", "make a post about", "content about", a Threads post, or an Instagram caption. For the full SEO/AEO/GEO blog-post workflow (keyword gap, answer-first structure, schema, internal linking, publishing), use blog-writing instead. For a direct offer or conversion goal, use copywriting-sales.
kindly-tools/skills/marketing/copywriting-awareness/SKILL.md
Write conversion-focused sales copy for offers, landing pages, ads, email sequences, and product pitches. Use when writing sales copy, writing an ad, writing a landing page, pitching a service, writing a cold email, writing a promo, closing the sale, or when the user says "write copy for", "sell this", "make this convert", "write an ad for", "sales email". Distinct from copywriting-awareness — use this when there's a clear offer and a desired action.
kindly-tools/skills/marketing/copywriting-sales/SKILL.md
Set up a new client analytics dashboard at {slug}.kindlycreativellc.com. Use when provisioning or scaffolding a client-facing analytics or reporting dashboard for a Kindly client.
kindly-tools/skills/marketing/content-pipeline--mktg/create-client-dashboard/SKILL.md
Deploy a new or existing website to production. Use when taking a site live, connecting a domain, or deploying any Kindly client/demo site to Vercel with a custom domain.
kindly-tools/skills/marketing/content-pipeline--mktg/deploy-site/SKILL.md
D2C product e-commerce conversion psychology — funnel anatomy, PDP structure, cart/checkout friction removal, mobile commerce mechanics, brand teardowns (Welly, Native, Allbirds, Quip, Meow Meow Tweet), and build-order argument for product brands selling physical goods. Companion to landing-page-psychology (which covers service-business homepages).
kindly-tools/skills/marketing/ecommerce-conversion/SKILL.md
Plan and write high-conversion email marketing. Use for email campaigns, broadcasts, newsletters, nurture sequences, welcome/cart/post-purchase flows, cold email, subject lines and preview text, deliverability, sender warm-up on new domains, drip sequences, segmentation, reply triage, offers, and seasonal send calendars. Covers copy tactics (Cialdini phrase banks), 80+ subject lines, template skeletons, voice rules, and pre-send checklists. For prospecting outreach tied to a specific researched lead plus a demo site (GoHighLevel 3-touch), use prospect-outreach instead. For SMS/text outreach use sms-outreach.
kindly-tools/skills/marketing/email-marketing/SKILL.md
High-converting forms and lead capture: field minimization, multi-step, validation UX, scheduling/booking, lead magnets, bot protection, speed-to-lead, abandonment recovery. Use when building or auditing any contact/booking/lead form.
kindly-tools/skills/marketing/forms-lead-capture/SKILL.md
Full content pipeline — write blog + post to Threads + post Instagram carousel/Reel. Use when asked to do a "full run", "content run", or "blog + social".
kindly-tools/skills/marketing/content-pipeline--mktg/full-content-run/SKILL.md
"Generative Engine Optimization (GEO): the EXECUTION work of getting cited by ChatGPT, Perplexity, Claude, and Gemini. Off-site citation signals (Reddit, earned media, reviews, Wikipedia), AI crawler access, llms.txt, content patterns LLMs cite, and Share-of-Voice measurement. Use when doing or improving AI answer visibility, AI referral traffic, off-site presence, or AI citation measurement. Do NOT use for upfront planning of what to write or target (ai-search-strategy owns the blueprint) or for on-page answer formatting (aeo-content-structure owns that)."
kindly-tools/skills/marketing/geo/SKILL.md
Full playbook for setting up, optimizing, and reporting on Google Ads campaigns for Kindly clients — campaign structure, keyword research, bidding, conversion tracking, and landing page requirements
kindly-tools/skills/marketing/google-ads/SKILL.md
GA4 setup, lead-gen conversion measurement (generate_lead/qualify_lead key events, phone + booking tracking, server-side conversions), funnel analysis, heatmaps/session replay, CRM attribution, consent-mode privacy, and reporting scripts for all Kindly client sites
kindly-tools/skills/marketing/google-analytics/SKILL.md
CRM setup, workflow automations, pipeline management, and API integration for GoHighLevel — covers lead response, nurture sequences, appointment booking, and review requests
kindly-tools/skills/marketing/highlevel/SKILL.md
"Create and post any format to Kindly Creative's Instagram: single image, carousel, or Reel. Covers the brand gate, copywriting, image hosting, and the full Instagram Graph API publish sequence for each format. Use when TJ says 'post this on instagram', 'post this screenshot', 'put this on IG', 'instagram this', 'make a carousel', 'create an IG carousel', 'make a reel', 'turn this blog into a reel', or sends an image to share. For posting to Threads, cross-platform cadence, token refresh, or engagement pulls, use social-posting instead."
kindly-tools/skills/marketing/instagram/SKILL.md
Conversion-focused homepage structure grounded in trust psychology and funnel logic. Load when building or critiquing any homepage or landing page — covers the temperature model, 9-section structure, dark/light visual rhythm, and conversion rules. Use when TJ says "landing page", "homepage structure", "why isn't this converting", "build me a homepage", or "funnel psychology". This is the persuasion, copy, and page-structure lane; for CRO methodology, benchmarks, A/B testing, form mechanics, and Core Web Vitals numbers, use conversion-optimization instead.
kindly-tools/skills/marketing/landing-page-psychology/SKILL.md
Local SEO and Google Business Profile: GBP optimization, LocalBusiness schema, citations, local/location landing pages, reviews, map pack, service-area + national hybrid. Use for any local or service-area business SEO.
kindly-tools/skills/marketing/local-seo/SKILL.md
Strategic-layer marketing reference for any Kindly Creative engagement. Covers positioning (Dunford framework), penetration-vs-loyalty (Sharp / Ehrenberg-Bass), funnel + customer journey design, awareness-level targeting (Schwartz), channel mix logic, brand vs. performance balance (Binet & Field 60/40), budget allocation, marketing diagnosis (where's the leak), roadmap sequencing for new clients, common strategic mistakes ranked, operating cadence, and the default Kindly approach for early-stage SMB clients. Load before any work that asks "what should we do" rather than "how do we do it" — campaign planning, channel selection, positioning workshops, funnel review, marketing roadmap for a new client.
kindly-tools/skills/marketing/marketing-strategy/SKILL.md
Full playbook for Facebook and Instagram advertising — campaign structure, audience targeting, creative best practices, Meta Pixel setup, A/B testing, and monthly reporting
kindly-tools/skills/marketing/meta-ads/SKILL.md
>- Craft irresistible offers that people feel stupid saying no to. Use this skill whenever the user wants to build, design, package, price, strengthen, critique, or rewrite an offer, promotion, deal, pitch, or value proposition, or turn one into sales/ad/landing/email copy. Trigger on phrases like "craft an offer," "make this irresistible," "my offer isn't converting," "package my service," "price my product," "write a sales page/VSL/ad," "value proposition," "what should I charge," "grand slam offer," or any request to make a product more compelling to buy. Works for any product, service, course, coaching, SaaS, agency, or local business. Use it even when the user doesn't say the word "offer" but clearly wants to make something more compelling to buy.
kindly-tools/skills/marketing/offer-architect/SKILL.md
>- Research a buyer demographic and generate a set of vivid, research-grounded customer personas (B2C consumers or B2B buyers) for use in offer crafting, website and landing-page copy, ad creation, copywriting, product, and sales. Each persona is a synthetic composite with demographics, finances, psychographics, behavior, goals, pain points, jobs-to-be-done, buying-decision insight, market weighting, and source/confidence flags. Use this skill whenever the user wants to understand their target audience, build buyer/customer personas, define an ideal customer profile (ICP), segment a market, figure out "who are we selling to," or needs personas to feed into the persona-panel skill for idea testing. Trigger even if the user just describes a product and asks "who would buy this" or "what's my target market" without saying the word "persona."
kindly-tools/skills/marketing/persona-builder/SKILL.md
>- Convene a panel of customer personas as AI judges that independently score and rank a marketing idea, then run a multi-agent QA/review process to make the ranking reliable. Use this skill whenever the user wants to pressure-test, grade, score, rank, or get feedback on an offer, headline, ad, landing page, email, sales pitch, pricing, product concept, name, or any piece of copy or creative — by asking "would my customers actually buy this?", "which version converts better?", "what would my audience think?", or "run this by my personas / my target market." It scores each idea on likelihood to convert, message clarity & relevance, believability & trust, and differentiation, aggregates the votes weighted by market share, surfaces objections, and recommends fixes. Pairs with the persona-builder skill (use it first if no personas exist yet). Trigger even when the user doesn't say "panel" — any request to evaluate an idea against an audience qualifies.
kindly-tools/skills/marketing/persona-panel/SKILL.md
Draft and post a single Threads post to @kindlycreativellc. Use when asked to post to Threads, write a Threads post, or share something on Threads. For auto-converting an existing blog post into a thread use social-posting (blog-to-social pipeline); for cadence, tokens, or engagement also use social-posting.
kindly-tools/skills/marketing/content-pipeline--mktg/post-threads/SKILL.md
Find and set up a new prospect. Use when TJ says to find, look up, or start a new prospect. Creates the folder structure, REPORT.md skeleton, and database entry.
kindly-tools/skills/marketing/prospect-find/SKILL.md
Write personalized cold outreach for a specific researched prospect in the prospecting pipeline. Use after the demo site is built and approved (follows prospect-find, prospect-research, prospect-site). Creates a 3-touch email sequence + SMS formatted for GoHighLevel. For general email campaigns, nurture sequences, or subject-line/deliverability work use email-marketing; for SMS strategy and templates use sms-outreach.
kindly-tools/skills/marketing/prospect-outreach/SKILL.md
Deep research and asset scraping for an existing prospect. Use after /prospect-find. Scrapes photos from Google Business, Facebook, and socials. Creates copy-samples.md with brand research and vibe assessment.
kindly-tools/skills/marketing/prospect-research/SKILL.md
Build and deploy a demo website for a prospect. Use after /prospect-research. Generates site-config.ts from research, builds the Next.js site from the template, and deploys to Vercel.
kindly-tools/skills/marketing/prospect-site/SKILL.md
Build, deploy, and verify a blog post to kindlycreativellc.com (the shipping step: MDX build, git commit/push, Vercel deploy, live-URL check, marketing-API pull). Use when shipping, publishing, or deploying a post to kindlycreativellc.com specifically. For the writing and SEO/AEO/GEO methodology (keyword research, structure, schema), use blog-writing first.
kindly-tools/skills/marketing/content-pipeline--mktg/publish-blog/SKILL.md
Pull latest analytics and content metrics from all platforms. Use when asked to check analytics, pull metrics, refresh data, or get performance numbers.
kindly-tools/skills/marketing/content-pipeline--mktg/pull-analytics/SKILL.md
Core SEO foundations, keyword research, on-page optimization (Yoast-grade checklist), off-page and link building, technical basics, rank tracking, and ranking timelines, current to mid-2026. Use when the user says "SEO", "optimize for search", "rank on Google", "keyword research", "meta title", "meta description", "on-page", "backlinks", "link building", "why isn't this page ranking", or asks how search works now. Routes to specialist skills: full site audits go to seo-aeo-audit, AI answer engines (ChatGPT, Perplexity, AI Overviews citations) go to aeo-content-structure and geo, FAQ/schema markup work goes to aeo-content-structure, Google Business Profile and Map Pack go to local-seo, writing blog posts goes to blog-writing, full keyword strategy blueprints go to ai-search-strategy (quick keyword questions stay here).
kindly-tools/skills/marketing/seo/SKILL.md
Five-layer site audit covering Technical SEO, On-Page Content, AEO (answer engines), GEO (generative engines / off-site citation), and Local SEO, with A-F grading, competitive analysis, and a client-facing report format for the AI-search era
kindly-tools/skills/marketing/seo-aeo-audit/SKILL.md
Writes high-response SMS and text-message outreach for sales, recruiting, and appointment booking, plus follow-up sequences, confirmations, and no-show protection. Evidence-ranked rules (speed, one binary ask, two slots, under 160 chars) from a 150-agent research compendium. Use when the user says "text campaign", "SMS outreach", "recruiting texts", "text the leads", "follow-up sequence", "get more replies", "book interviews by text", "reduce no-shows", "GHL texting", or any task writing texts meant to get a reply or booking. Includes a deployable AI texting-agent kit (prompt, knowledge base, GHL build spec, test rubric). Do NOT use for email outreach (email-marketing-playbook) or one-off transactional notifications.
kindly-tools/skills/marketing/sms-outreach/SKILL.md
"Cross-platform social ops for Kindly Creative: Meta token management (refresh 1st and 15th), content calendar cadence and content mix, blog-to-social pipeline, and engagement metric pulls. Use for 'refresh the social tokens', 'what's our posting cadence', 'plan the content calendar', 'turn this blog into a thread', or 'pull engagement'. For actually posting a Threads post use post-threads; for creating or posting Instagram content use the instagram skill."
kindly-tools/skills/marketing/social-posting/SKILL.md
Produce on-brand animated marketing content (kinetic typography, animated logos and reveals, motion graphics, social video clips, overlays, data animations) using Swishy (swishy.ai), the AI motion-design tool that turns a text prompt into an exported MP4/MOV/GIF in about 30 seconds. Use when the user says "use swishy", "make a swishy animation", "swishy this", "animate this text", "kinetic typography", "animated logo", "logo reveal", "motion graphics for a reel", "animate our logo for the reel", "make a text animation for instagram/tiktok", or wants a quick animated clip, hook, title, or motion overlay for social or ads. Covers the curated map of Swishy templates to Kindly content needs, on-brand prompting, aspect ratios, and the brand plus approval gate before posting. Do NOT use for on-site website or UI motion in code (use web-animation, that is CSS/Lottie/Rive rendered on the page, not exported video), for editing existing footage (use video-creation or remotion), or for static social images and carousels (use instagram or kindly-design).
kindly-tools/skills/marketing/swishy-motion/SKILL.md
Trust, social proof, and credibility systems: testimonials, case studies, reviews, guarantees, authority/E-E-A-T signals, quantified results. Use when designing proof sections, choosing trust signals, or boosting credibility on a page.
kindly-tools/skills/marketing/trust-social-proof/SKILL.md
Systematic UI/UX review checklist covering attention flow, CTA prominence, visual design, typography, color, micro-interactions and loading states, form UX, and WCAG 2.2 AA accessibility, always reviewed at both mobile (390px) and desktop (1280px)
kindly-tools/skills/marketing/ui-ux-review/SKILL.md
Harry Dry's 3-question test for writing headlines, taglines, slogans, hero lines, ad hooks, and value props that stick in memory and that a competitor literally cannot copy. Run every line through three questions: Can I visualize it? Can I falsify it? Can nobody else say it? Three yeses means you are onto something, three nos means it is generic filler. Includes the scoring rule, the worked New Balance and University of East London examples, fix levers to turn each No into a Yes, headline patterns that pass, a before/after gallery for local-service and e-commerce clients, and Harry Dry's supporting principles (show don't tell, be specific, echo the customer). Use when writing or auditing any short high-stakes line, or whenever copy feels vague, generic, swappable, or forgettable. Pairs with copywriting (the mechanics layer), copywriting-sales, offer-architect, and landing-page-psychology.
kindly-tools/skills/marketing/unforgettable-copy/SKILL.md
Two-pipeline video production playbook — Remotion for programmatic animated videos and HeyGen for AI avatar blog-to-video conversions — with post-production optimization and upload
kindly-tools/skills/marketing/video-creation/SKILL.md
Full playbook for building Kindly client websites (Next.js 16 App Router + Sanity + Vercel stack), rendering and caching strategy, headless/structured CMS modeling, staging/preview workflow, speed-to-lead infra, security baseline, Core Web Vitals (LCP/INP/CLS 2026 thresholds), WCAG 2.2 AA accessibility, SEO/AEO checklist, and pre-launch verification
kindly-tools/skills/marketing/website-building/SKILL.md
Strict process for making and verifying changes to client websites — baseline screenshots before edits, visual QA after, Telegram approval from TJ, then commit and deploy
kindly-tools/skills/marketing/website-changes/SKILL.md
Visual design baseline for every Kindly Creative client website, layout, typography, color rhythm, glass cards, bento grids, motion, photography, imagery that converts, color/contrast for CTAs, visual hierarchy, micro-interactions, mobile-first, navigation, and iconography. Brand-agnostic; bring the client's own colors and fonts on top.
kindly-tools/skills/marketing/website-design/SKILL.md
Draft Rana Golf blog posts and on-site product/collection copy, brand-voice and kill-word checked, with SEO keyword targets baked in. Use for blog drafts, product descriptions, collection copy, and on-page SEO content.
kindly-clients/rana-golf/.claude/skills/content-blog/SKILL.md
"Drafts, builds, and stages Rana Golf marketing emails and SMS in Klaviyo, then carries a campaign all the way to a hard-gated send or schedule. Walks through audience, offer or purpose, template build, and campaign creation. Use whenever the user mentions an 'email campaign', 'Klaviyo campaign', 'email blast', 'draft an email', 'send an email to the list', 'brand story email', 'promo email', 'welcome email', 'abandoned cart', a flow, a segment, or blasting the Email List, even if they do not name this skill. Every send or schedule stops for TJ's explicit per-action yes. Do NOT use for blog or product copy (content-blog), social posts or carousels (social-carousel), or performance reports (performance-report)."
kindly-clients/rana-golf/.claude/skills/email-campaign/SKILL.md
"Write or edit any first-person Micah piece for Rana Golf (campaign emails, founder notes, B2B club outreach, review replies, SMS in his voice) so it reads like Micah actually wrote it. Built from his founder questionnaire (2026-06-28), his copy feedback, and a corpus of 76 emails he sent March 2025 to June 2026. Use whenever drafting or revising founder-voice copy, whenever the user says 'make it sound like Micah', 'Micah voice', 'would Micah like this', or before any email copy goes to Micah for review. Complements (does not replace) context/brand-voice.md and the brand checklist; this skill is the how-he-actually-writes layer."
kindly-clients/rana-golf/.claude/skills/micah-voice/SKILL.md
Pull Rana Golf performance data from Shopify, Klaviyo, and GA4 and produce a periodic marketing performance report. Use for monthly/weekly reporting, campaign post-mortems, and KPI summaries.
kindly-clients/rana-golf/.claude/skills/performance-report/SKILL.md
Plan and draft Rana Golf organic social posts and carousels for Instagram, TikTok, and Facebook from the content calendar. Use for post captions, carousel frame plans, and content pillars. Drafts only.
kindly-clients/rana-golf/.claude/skills/social-carousel/SKILL.md
Keep a live room engaged - wake a sleepy section, recover a flat moment, decide where the audience should interact, make a slide land harder, or diagnose why energy dropped. Grounded in cognitive-load theory, Mayer's multimedia principles, and attention research. Use for live-room energy, participation, and engagement diagnosis. For talk structure use speech-writing; for hooks and story use storytelling-hooks; for delivery cue cards use presenter-notes. argument-hint: '[talk, slide, or moment to energize]' allowed-tools: - 'Read' - 'Write' - 'Edit' - 'Glob' - 'Grep'
kindly-tools/skills/training/audience-engagement/SKILL.md
Produce condensed LIVE-DELIVERY cue cards (cues, not scripts) that drive spoken delivery from a deck - the confidence-monitor notes. Use for "write speaker notes", "presenter notes for this slide", "what should I have on the confidence monitor", "what do I say between these slides". For writing the full talk script and structure, use speech-writing; for openers/closers and story, use storytelling-hooks; for live-room energy, use audience-engagement. argument-hint: '[slide or section needing notes]' allowed-tools: - 'Read' - 'Write' - 'Edit' - 'Glob' - 'Grep'
kindly-tools/skills/training/presenter-notes/SKILL.md
Write or fix the STRUCTURE and SCRIPT of a talk - the spine, three-act arc, sparkline, transitions between slides, pacing, and word economy. Use for "write a talk", "draft a keynote", "structure a presentation", "fix the flow of this speech", "tighten this script", "what should I say on this slide", talk tracks. For openers, closers, hooks, and narrative, use storytelling-hooks; for condensed live-delivery cue cards (confidence monitor), use presenter-notes; for live-room energy and participation, use audience-engagement. argument-hint: '[topic, talk, or slide to write/refine]' allowed-tools: - 'Read' - 'Write' - 'Edit' - 'Glob' - 'Grep'
kindly-tools/skills/training/speech-writing/SKILL.md
Find or build the STORY in a talk - hooks, openers, callbacks, narrative spine, and closes. Use when "the hook isn't landing", a section "feels flat", a draft reads like a list of facts, or for "how do I open this" / "how do I close this". For overall talk structure and flow use speech-writing; for live-delivery cue cards use presenter-notes; for live-room energy use audience-engagement. argument-hint: '[talk, section, or slide that needs a story or hook]' allowed-tools: - 'Read' - 'Write' - 'Edit' - 'Glob' - 'Grep'
kindly-tools/skills/training/storytelling-hooks/SKILL.md
Use when the user wants to design a training that sticks, plan a workshop or onboarding session, build a training program, fix forgotten content ("people forget what I taught"), structure how to teach a topic, or asks "how do I teach X so it sticks" or "make this stick". argument-hint: '[topic or training to design]' allowed-tools: - 'Read' - 'Write' - 'Edit' - 'Glob' - 'Grep'
kindly-tools/skills/training/training-design/SKILL.md
Web accessibility and WCAG 2.2 / ADA compliance baseline: semantic HTML, keyboard, focus, contrast, forms, aria, testing. Use when auditing or building for accessibility, reducing ADA legal risk, or improving usability for everyone.
kindly-tools/skills/web-dev/accessibility/SKILL.md
Design the interaction model and visual taste for an agent-native app, separate conversation thread from a durable activity/run timeline, transparency-first tool disclosure, tiered human-in-the-loop approval cards as structured logged replayable events, plan-and-execute step views, presence and multi-agent threads, and Refactoring-UI plus Apple-HIG craft (grayscale-first hierarchy, constrained scales, coarse status, alert discipline). Use when designing any approval, status, presence, or rich-content surface.
kindly-tools/skills/web-dev/kindly-comms/agent-app-interaction-design/SKILL.md
How to use Blender from Claude via the BlenderMCP server. Covers install + addon + .mcp.json wiring, the available MCP tools, scene-authoring patterns (geometry, lighting, materials, animation), the render pipeline (PNG sequence → ffmpeg → MP4/WebM/WebP), and the gotchas that bite in Blender 5.x. Load when TJ says "blender", "render in blender", "3D scene", "voxel scene", "hero video loop", "animate in 3D", "blender mcp", or "addon.py" — or when a task needs a 3D asset rendered to ship.
kindly-tools/skills/web-dev/blender-3d/SKILL.md
Design and audit one backend API/contract that serves the web PWA and the later native iOS app as interchangeable consumers, transport-neutral typed payloads, auth split by client capability, first-message or header WebSocket auth, additive-only versioning, cursor pagination, a stable coded error envelope, and media-by-reference with Range support. Use when adding any endpoint, WebSocket event, or shared type, or auditing whether a surface is PWA-only.
kindly-tools/skills/web-dev/kindly-comms/client-agnostic-api-design/SKILL.md
Reverse-engineer and clone one or more websites in one shot — extracts assets, CSS, and content section-by-section and proactively dispatches parallel builder agents in worktrees as it goes. Use this whenever the user wants to clone, replicate, rebuild, reverse-engineer, or copy any website. Also triggers on phrases like "make a copy of this site", "rebuild this page", "pixel-perfect clone". Provide one or more target URLs as arguments. argument-hint: "<url1> [<url2> ...]" user-invocable: true
kindly-tools/skills/web-dev/clone-website/SKILL.md
Playbook for deploying Kindly client sites to Vercel with Cloudflare DNS — covers standard deploys, env vars, media uploads, rollback, and multi-machine sync
kindly-tools/skills/web-dev/deployment/SKILL.md
Gives an agent a taste layer for anything visual meant to impress a human, landing pages, heroes, animation concepts, posters, social visuals. Names what reads as genuinely cool to a human eye in 2026 versus templated AI-slop, four registers to choose and commit to, a named type menu, a kill list of current AI tells, and a decision procedure for naming real references and holding one point of view. Use when making or reviewing a landing page, hero, animation or motion concept, poster, or social visual, or when the user says "make it cool", "make it interesting", "not boring", "this looks AI-generated", "does this look like a template", "AI-slop check", "give this personality", or "make it stand out". Composes with website-design, web-animation, ui-ux-review, and landing-page-psychology, it does not replace them. Do NOT use for usability bugs, WCAG compliance, motion implementation (web-animation's job), or copy and funnel-structure questions with no aesthetic angle.
kindly-tools/skills/web-dev/front-end-human-eye/SKILL.md
Architect a maintainable native iOS (Swift/SwiftUI, iOS 17+) real-time agent chat client, Observation (@Observable) state, typed-Route NavigationStack with a Router, async/await URLSession behind a protocol with an actor for token refresh, @Environment dependency injection, SPM module boundaries, GRDB/SQLiteData over SwiftData, Swift Testing, accessibility in CI, and a native smoke test mirroring bun run smoke. Use when the iOS spec is opened or any native-client decision is made.
kindly-tools/skills/web-dev/kindly-comms/ios-swiftui-architecture/SKILL.md
Default to the simplest thing that works when writing or dispatching code, and aggressively avoid over-engineering. Apply when building or changing code, adding a dependency or library, scaffolding a feature, or whenever a build is growing more complex than the problem it solves. Trigger phrases include "don't over-engineer this", "keep it simple", "the simplest thing that works", "do we really need that dependency", "make it lean", "lazy senior dev", "the best code is the code you never wrote". Inspired by Ponytail (github.com/DietrichGebert/ponytail).
kindly-tools/skills/web-dev/minimal-code/SKILL.md
Visual QA workflow using Playwright and the browser tool — screenshots at mobile (390px) and desktop (1280px), accessibility checks, Lighthouse targets, and common issue fixes
kindly-tools/skills/web-dev/playwright-testing/SKILL.md
Design reliable real-time messaging that produces identical state across the web PWA and a native iOS client, WebSocket reconnect with backoff and jitter, an app-level heartbeat, resume/replay on every reconnect, per-conversation seq ordering, at-least-once delivery with idempotent send and receive, optimistic UI reconciliation, durable offline outboxes, and push as a wake-only nudge over one budget. Use when specifying connection lifecycle, delivery guarantees, or notifications.
kindly-tools/skills/web-dev/kindly-comms/realtime-message-sync/SKILL.md
Playbook for creating programmatic videos with React/Remotion — compositions, animation patterns, data-driven rendering, and upload to Vercel Blob
kindly-tools/skills/web-dev/remotion/SKILL.md
Canonical folder layout and required files for every Kindly Creative repo — the 3-zone model, skill-vs-workflow rule, full client-repo spec with context/, integrations/, data/, memory/, workflows/, and the root-playbook boot model
kindly-tools/skills/web-dev/repo-structure/SKILL.md
Render untrusted agent-authored HTML/CSS/JS safely as artifacts and interactive blocks, separate-origin sandbox iframe (allow-scripts without allow-same-origin), per-artifact CSP, a validated postMessage/JSON-RPC bridge, a capability-scoped host API funneled through the permission relay, the MCP-Apps/SEP-1865 predeclared ui:// pattern, a safe inbound interaction-to-agent notification path, and the matching WKWebView boundary on iOS. Use when designing the artifact playground or any block that runs or posts back agent code.
kindly-tools/skills/web-dev/kindly-comms/secure-agent-authored-ui/SKILL.md
Headless Shopify storefront playbook using Next.js and the Storefront GraphQL API: project setup, Storefront queries, Cart API + hosted-checkout flow, ISR, and performance. Use for building a custom React/Next.js storefront against Shopify. For onboarding a client onto a Shopify Liquid THEME (template scaffolding interview), use shopify-storefront-onboard instead.
kindly-tools/skills/web-dev/shopify/SKILL.md
Onboard a new client onto the Shopify Liquid theme template (kindly-templates/shopify-liquid-storefront) - a structured interview that fills brand tokens, fonts, voice, ICP, value props, offerings, and integrations, then builds and commits the client repo. Use when scaffolding or onboarding a new Shopify Liquid THEME client from a fresh template clone ("onboard this client", "scaffold a new client", "run the shopify onboarding"). For building a headless Next.js + Storefront API storefront, use shopify instead.
kindly-tools/skills/web-dev/shopify-storefront-onboard/SKILL.md
Apply transferable software-design mental models (complexity-as-enemy, simple-vs-easy, deep modules, define-errors-out-of-existence, design-it-twice, AHA over DRY) to make and defend kindly-comms architecture decisions and police module seams. Use before writing or reviewing any spec, when naming an abstraction, when a change touches many files, or when tempted to add a config knob, a layer, or a clever generalization.
kindly-tools/skills/web-dev/kindly-comms/software-design-judgment/SKILL.md
Set up Cloudflare Turnstile bot protection on a Kindly client site's forms, end to end (widget creation, Vercel env wiring, code integration, verification, fail-closed hardening). Use when the user says "add turnstile", "set up turnstile for <client>", "the contact form is getting spam", "add bot protection", "add a captcha", or when a form audit flags missing spam protection. Also the reference for VERIFYING an existing Turnstile install (widget renders, enforcement active, real users pass).
kindly-tools/skills/web-dev/turnstile-bot-protection/SKILL.md
Reference for designing and shipping web animation on Kindly client sites. Covers the 2026 trend snapshot (voxel/pixel revival, synthwave, scroll-driven, hero loops), the library decision tree (CSS vs Lottie vs Rive vs GSAP/ScrollTrigger vs Framer Motion vs WebGL), Stripe-school performance principles, accessibility rules, and the Kindly Creative locomotive motion system. Use when TJ says "add animation", "hero animation", "scroll animation", "motion design", "animate this", "hero video loop", "Lottie", "Rive", "GSAP", "retro animation", "pixel art animation", "voxel scene", or "make it feel alive".
kindly-tools/skills/web-dev/web-animation/SKILL.md
Web performance and Core Web Vitals for conversion and SEO: LCP/INP/CLS, image/font optimization, JS reduction, caching/CDN, Next.js App Router perf, mobile performance. Use when a site is slow, failing CWV, or before launch.
kindly-tools/skills/web-dev/web-performance/SKILL.md