550 skills. 31 hubs.

Build AI features and apps on the major model providers (OpenAI, Anthropic/Claude, and Google Gemini) the model-agnostic way, so you can swap providers without rewriting.

Get the real images, icons, illustrations, fonts and footage a build needs, decide source versus generate versus commission, pull from licensed libraries, verify every licence and attribution requirement, treat assets into the palette, and record the provenance so it ships with the work. Owns sourcing and provenance; for art direction of generated imagery use as-design-image-styles.

Your on-demand branding consultant: get expert, reasoned answers to any brand DECISION instead of a one-shot kit.

From a brand name + 3-5 vibe words, produce a complete mini brand book (positioning one-liner, color palette + type pairing, logo/mark concepts, and a voice mini-guide) assembled into one client-ready package (a $1-5k starter deliverable).

Work a single component to a finished state, generate real variants, prototype the interaction, then deliberately break it against long content, empty data, slow networks, errors, narrow widths, keyboard-only use and reduced motion. Returns the component plus the evidence it survives.

Produce concept and editorial illustration for a build, decide illustration versus photograph versus diagram, hold one visual system across a whole set, recolour and adapt licensed illustration into the palette, and keep a reusable library with its provenance. Owns illustration as a system; for photographic generation use as-design-image-styles, for explanatory schematics use as-design-diagram.

Give structured, severity-tagged design feedback on a screen, mockup, or built component, usability, hierarchy, consistency (with the a11y lens routed to the WCAG gate). Critiques designs against the Agency Script system; for building the UI use as-ops-premium-ui-builder, for the WCAG gate use as-core-app-responsive-a11y.

Build a reusable visual theme for the things you hand over (decks, reports, proposals, one-pagers, posters, certificates, covers) then compose and render pieces from it through whichever design tool is actually connected (Canva, Adobe Express, Gamma, an HTML pipeline, or a document toolchain).

Create an editorial-grade diagram (architecture, flowchart, sequence, state machine, ER/data model, swimlane, org chart, timeline, Sankey, Wardley map, user journey, deployment, dependency graph, UML class, database schema, and 24 more types) as a self-contained branded HTML/SVG figure, or redraw an existing draw.io/Mermaid diagram into presentable form.

Translate Figma designs into production-ready front-end code with high visual fidelity, pull design context, screenshots, variables/tokens, and assets via the Figma MCP, map them to the app's real design system (Tailwind/shadcn/tokens, not hardcoded values), and build accessible, responsive components that match the design 1:1.

Generate a developer-handoff spec from a design (layout, exact tokens, component props/API, interaction states, responsive breakpoints, edge cases, and motion) so engineering builds it once, correctly. Produces the buildable spec; for the WCAG gate use as-core-app-responsive-a11y, for building it use as-ops-premium-ui-builder.

Identity-preserving, visual-first portrait + photo-presentation studio, professional/personal-brand headshots (LinkedIn, founder, speaker, author, press, bio), tasteful social/dating profile photos, photo cleanup, and infographic presentation reports (face, hairstyle, outfit/color, grooming, profile-photo) with improved same-person previews and presentation-only scores.

Design category hub: routes to the right leaf for logo, brand, merch, diagram, visual-modes, visual-prompt-library, photoshoot, QR, theme, critique. Entry point for the design domain.

Apply print and retro treatments to an image you already have, 16-bit sprite, fat pixel, bootleg print, ASCII, green terminal, risograph, 1-bit Atkinson dither, punk collage, CMYK rosette dots, halftone print. Deterministic pixel math on our own engine, so composition is preserved by construction, byte-for-byte reproducible, $0 per image at any batch size.

Run a standing intake that turns what you see into what you can build, capture references daily with the read attached, file them by structural dimension rather than by vibe, and periodically build a small deliberate exercise against one captured idea so the input becomes capability. Pairs with as-design-reference-mining, which mines one reference for one build.

Run one cross-discipline review of a screen, flow, or interface change and return a single ranked verdict, accessibility, layout, hierarchy, density, affordance, typography, colour, polish and motion, each finding pinned to a location, cited to a named law, and priced by user impact.

Turn a system, process, Brain section, or pile of notes into a portable, diff-able visual graph (a valid JSON Canvas (.canvas) file (jsoncanvas.org 1.0 open spec) nodes[] + edges[]) laid out as a mind-map, flow, or architecture diagram, grounded in real source content.

Logo and the full identity system, mark, wordmark, lockups, favicon, one-colour, reversed. Reads the live site first.

Author a whole expert method as a sellable Mega Prompt (role, anti-roles, phases, gates, rubric, refuse list) then hand it to as-design-megaprompt-compiler to encode as versioned JSON. One-prompt rewrites stay as-design-prompt-engineering; encoding an already-written method stays the compiler; chaining different skills stays as-core-workflow-builder.

Encode an already-authored expert method as a versioned JSON MegaPrompt spec, schema-validated, compiled per model (Claude first), inference-traced, rubric-scored. Ends prompt drift.

Design merch your audience will actually wear (T-shirts, hoodies, hats, stickers, mugs) that look professional, not amateur print-on-demand.

Art-direct a commissioned AI photoshoot end to end, classify the use case (profile picture, print magazine, video ad, commercial, UGC, e-commerce, hero, billboard, thumbnail, food, interior, event), run a full seven-round creative intake, lock a signed-off Shoot Spec Sheet, then write model-ready prompts with camera, lens, aperture, lighting pattern, ratio, palette, and grade specified.

Data-driven bulk image rendering: feed a CMS table or CSV, emit hundreds of on-brand OG cards, ad variants, and listing images from parameterized templates, token-driven and QA'd at scale. Single bespoke assets go to as-design-system / as-ads-meta-creative; this is the at-scale line.

Write, optimize, and test a prompt that works on any model, the general text-prompt craft, distinct from authoring a skill or an image prompt.

Generate branded, scannable QR codes that people actually want to scan. Not boring black-and-white squares. Produces an on-brand QR (brand colors with safe contrast, a center logo, custom eyes, rounded modules) at high error-correction so it stays reliable, with a tracked/short destination URL, exported print-ready (SVG + high-res PNG) at the right quiet-zone and minimum size.

Add real-time graphics to a web page on purpose (ambient grounds, pointer-response fields, 3D objects and materials, globes, environmental scenes, shader effects and physics) each admitted through a four-part test, priced for GPU and weight, shipped with a static fallback and a reduced-motion path.

Make a UI codebase visually consistent and KEEP it that way, audit the component tree for drift (distinct CTA treatments, hardcoded hex, arbitrary radii/type/z-index, nav sprawl), consolidate one primitive at a time into token-backed components, and install a drift ratchet so the numbers only go down.

Turn a reference (a screenshot, a Pinterest board, a live site, a video, or a full-page capture) into a measured design brief carrying an extracted palette with a real contrast matrix, density and rhythm read, corner and depth language, type register, transcribed motion, and a direction contract. Every token anchored to something visible in the reference.

Run the surface-reduction pass on a real app, inventory every screen, nav item, action, field and setting; name one job per screen; measure decisions/steps/fields to value on the top tasks; collapse the information architecture; default instead of asking; defer power to advanced surfaces with a findable path; then publish before/after counts.

Crawl a whole website or web app page by page in a real Chromium browser, capture full-page desktop/mobile screenshots plus DevTools/DOM evidence, force an individual visual review of every captured route and safe UI state, and synthesize one prioritized design-system, UI/UX consistency, responsive, accessibility, and AI-slop audit.

Audit a site, app, screenshot, mockup, or UI codebase for the tells that it was generated rather than made, placeholder copy and names, round fake numbers, filler superlatives, default framework palettes, three-column feature-card grids, flat single-layer shadows, gradient-blob heroes, and reference material carried across without originality. Returns a scored, located, fixable list.

Audit, document, and extend the app's design system (tokens, component variants/states, naming consistency, and hardcoded-value drift) grounded in brand-config/Tailwind/shadcn. Owns design-system documentation + token hygiene; for building premium UI use as-ops-premium-ui-builder, for the WCAG gate use as-core-app-responsive-a11y.

From a product type + audience + stack, select and FREEZE a coherent, industry-grounded design system (style, palette, typography, landing pattern) to a MASTER spec (with per-page overrides) that every downstream builder binds to. Upstream of as-ops-premium-ui-builder.

Plan, run, and synthesize user research for the product, research plans, interview guides, surveys, usability tests, and synthesis into themes/insights/segments tied to a design decision. Owns product/UX research; for sales account research use as-sales-account-research, for market analysis use as-ops-competitor-intel.

Write and review in-product UX microcopy for the Next.js app (buttons, labels, error messages, empty states, CTAs, onboarding, toasts, tooltips) in the Agency Script voice. For marketing/landing copy use as-funnel-copy; for whole-app voice rewrites use as-core-brand-voice-guard; this owns interface strings.

Compile visual shot-language codes into original prompts and produce the picture, exploded views, cyanoprint engineering drawings, turnaround sheets, cutaways, macros, sketchnotes, knolling, schematics, storyboards, and the rest of the mode catalog.

Generate an original, on-brand, reusable VISUAL PROMPT PACK / library for AI image + video generation, master prompts across the styles a brand needs (photography, illustration, 3D/render, avatars, logos/UI graphics, Sora/Veo motion) plus a shot-language mode deck (exploded, cyanoprint, turnaround, knolling, …), each with the brand palette pinned, swappable subject slots.

The Writing Engine front door. Write ANY marketing asset on a proven, named framework, governed end to end. Pick the asset (sales page, social, email, ad, VSL, script, bio, offer), answer the audience, brand and offer questions once, then iterate by plain command ("funnier / more premium / shorter").

Rewrite flat sales copy into immersive, high-conversion prose, the sentence-level craft (arresting openings, open loops, recognition mirrors, future pacing, objection steel-manning, rhythm engineering) that turns a page with all the right sections into one people actually finish. Diagnoses why copy reads flat, fixes it line by line, and owns the downsell and thank-you surfaces.

Turn a proven AI workflow into money two ways, a done-for-you AI service (charge premium per project) or a productized AI tool/bot sold on a Stripe subscription or one-time fee ("sell the system, not the work"). Encodes the workflow into a well-instructed, knowledge-grounded bot, gates it with a free trial → Stripe billing, wraps a sales page + ICP, prices it on value.

Build an interactive SCORED assessment / readiness scorecard end to end, a multi-dimension diagnostic that gives the taker an overall score, a tier, and a personalized report, and gives you a qualified, segmented, scored lead.

Analyze a target audience and produce a deep, voice-of-customer customer avatar (the pains, desires, objections, beliefs, buying triggers, and the EXACT words they use) plus their awareness level and sophistication stage, so every piece of copy is written in their language at the right entry point.

Map one offer's messaging across all five awareness stages at once (unaware -> problem -> solution -> product -> most-aware) (the lead, headline angle, proof load, CTA, asset, and channel for each rung) so the whole funnel is matched stage by stage and walks a market up the ladder. Plans the full ladder, not the single dominant stage.

Build high-intent "[you] vs [competitor]" and "best [competitor] alternatives" pages that win bottom-funnel SEO and double as sales enablement. Honest by design (fair comparison table, an explicit "choose them when…", your proven wins, a migration angle, switcher proof, schema, and a CTA) every competitor fact true, current, and sourced (no disparagement, no fabrication, trademark-careful).

Design a high-ticket physical + install order flow: configure, deposit, slot, delivery, referral. Digital OTOs stay as-funnel-upsell-architect.

Design a recurring-revenue offer (membership, subscription, retainer, or continuity-on-a-core) built to RENEW, not just sign up. Picks the model, defines the recurring value unit ("what they get this month"), tiers it by outcome, prices it for LTV (monthly vs annual, founding rate, unit economics that clear CAC), engineers honest stickiness, plans the value calendar, and specs the retention loop.

Write or refresh sales pages, pricing pages, checkout, and success pages for Agency Script offers.

Find conversion leaks and design prioritized A/B tests across the funnel.

Spec the free-interactive-tool growth motion (engineering-as-marketing), a tool-selection matrix over calculator, generator, grader, template, and mini-app; SEO-keyword-to-tool fit; an ROI build-vs-value gate so you never build a tool the traffic won't justify; shareability and virality mechanics; and the tool-to-product path into lead capture and the paid offer.

Design growth loops (flywheels) where the output of one cycle becomes the input of the next, so growth compounds instead of needing ever-more spend, unlike a linear, leaky funnel. Names the four beats (input → action → output → reinvestment), confirms the reinvestment actually closes the loop, quantifies the per-beat conversion factors to find the weakest beat, fixes it, and shortens loop speed.

The product-marketing strategy deliverable for any client, a positioning statement built bottom-up (alternatives, attributes, value, who-cares ICP, market frame), a one-page message house (roof, pillars, foundation), per-segment messaging, and a GTM-motion pick by rubric. Upstream of the offer and the copy; every proof pillar traces to evidence or is cut.

Design and build a complete marketing/sales funnel end to end. Pick the right proven funnel map for the offer, then generate every page, email sequence, automation, analytics hookup, and monetization step.

Build complete lead magnets (guides, audits, quizzes/assessments) plus the opt-in page, delivery emails, and segmentation.

Apply behavioral science to a page, offer, email, pricing, or flow, the working library of principles (social proof, loss aversion, anchoring, scarcity, reciprocity, commitment, authority, framing, cognitive ease, endowment, peak-end, defaults, decoy) mapped to exactly where each applies.

Find a profitable niche fast and ideate the digital product to serve it, a guided 4-stage engine (Value Machine → unfair-advantage or business-advantage discovery → the 6-Factor Market Test scored out of 60 → niche lock-in + a 30-Day Outlook → 4-6 fast 7-14-day product concepts with outcome-driven names + WITHOUT hooks).

Turn a short 6-field intake (name, email, offer idea, deliverables, statement of value, 3 objections) (or an offer doc) into a detailed customer avatar plus 3+ high-ticket offer concepts, each value-stacked, premium-priced, guarantee-backed, and built to dissolve the objections. The mega-prompt offer engine.

Reverse-engineer the exact person you want on sales calls into one linked acquisition system, enforces the avatar-first gate (requests worksheet + call transcripts first), sequences the symptom→problem→root hierarchy, turns it into a problem-linked offer, splinters micro-offers/lead magnets.

Design and optimize the in-product monetization moment for product-led growth, paywalls, upgrade screens, upsell modals, feature gates, usage-limit prompts, and trial-end surfaces. Hard no-dark-patterns gate. No confirmshaming, fake countdowns, trick defaults, or cancellation mazes.

The empirical layer under pricing decisions, willingness-to-pay research run properly (van Westendorp, Gabor-Granger, simplified conjoint), trust-safe price/packaging test design, value-metric selection (per-seat vs usage vs outcome), elasticity from own history, discount governance and approval ladders, and price-change migrations with grandfathering math and churn-risk modeling.

Package a program or software offer into a three-rung delivery ladder (DIY, Done-With-You, Done-For-You) with defensible price anchors, a recurring platform fee underneath, and a fulfillment plan per rung, separating the asset (recurring) from the transformation (one-time) and use from resell. The packaging engine for program creators.

Productize a coaching/consulting/agency service into clear, tiered, priced packages (using the DSP framework (Desire → Solution → Package) and the MHI pricing math (income goal ÷ clients = base price; income goal ÷ hours = min hourly rate)) plus an action price, value-stacking bonuses, and a defensible floor. Delivers a good/better/best ladder and a clean packages table.

Lift average order value on a live funnel by engineering the post-purchase monetization stack (the order bump, one-click OTOs (one-time offers), the downsell that recovers a no, and the value ladder they hang on) each a genuine next step that helps the buyer reach the outcome faster, priced honestly, with no forced continuity.

Write a full long-form VIDEO SALES LETTER script (the 8-30 minute direct-response video that carries a VSL funnel) using the proven 12-beat structure (pattern interrupt + promise → call-out → story → reframed enemy → unique mechanism → proof → offer → value stack → guarantee → objection demolition → single CTA on a loop).

One input (a landing-page/funnel URL) → a complete, branded, client-ready CRO teardown report, scrape the page + flow, score it against the funnel/CRO canon, diagnose the conversion leaks, and hand back prioritized A/B hypotheses with an annotated, section-by-section teardown. The sellable lead magnet, end to end.

Act as an AI Film Director, turn a rough idea, script, ad, or character concept into a render-ready cinematic production package for image+text AI video models (Veo3, Higgsfield, CoreReflex/WAN, and others).

Take a 9:16 talking-head video and overlay animated titles, step numbers, keyword badges, progress bars, and synced captions (Remotion/Hyperframe), original video stays full-frame.

Generate a 3-6s animated logo sting / brand intro or outro (Remotion/Hyperframe), logo reveal with spring scale, particle burst, tagline, brand-color glow.

Turn a CSV (or pasted numbers) into an animated data-dashboard video (Remotion/Hyperframe), a hero KPI count-up, growing bar chart, self-drawing donut, and animated line chart.

Edit REAL recorded footage into a finished video, transcribe multi-take talking-head clips, pick the best take per scene, cut filler/warm-ups on word timing, stitch with FFmpeg, color grade, and add Remotion title/caption frames. The real-footage editing path (vs. as-video-hub which generates video from code).

Turn any topic into a fully animated explainer video (Remotion/Hyperframe), research, script, design, animate.

Produce a branded, character-led animated explainer end to end on the CoreReflex API, entity card intake, Format-N script, storyboard approval gate, one reused style token, illustrated plates + self-hosted VO + music bed, free engine retiming, verified render, and the measured case study. The character, palette and mission swap per entity; the pipeline is fixed.

Turn one topic or course module into a full multi-video explainer series, segmented outline, per-segment scripts, one shared design system, batch-rendered animated videos (via as-video-education-explainer-video), assembled into a playlist with thumbnails.

Produce a Format 1 Honest-Version explainer storyboard (7 chapters, 21 beats (HOOK / FACT / TURN / RULE), native-shape cards, sourced claims gate) as a self-contained HTML artifact and a CoreReflex longform composition brief. Not the 30-second Remotion explainer (as-video-education-explainer-video) and not cinematic AI video (as-video-ai-director).

Natural-language editing intent into deterministic FFmpeg pipelines (reframe to 9:16/1:1/16:9, burn captions, add a ducked music bed, trim, concat, transcode per platform, loudness-normalize, extract stills/GIFs) on existing footage, source preserved, output probe-verified.

The video front door (routes between EDITING real footage (as-video-edit-multitake) and GENERATING animated video from code (Remotion / Hyperframe)) explainers, product demos, testimonials, avatar overlays, data dashboards, social shorts, brand intros. Entry point for the video domain.

Shot bible and banned tropes for factory / test-stand / line film so industrial clients do not look like SaaS ads. Briefing is as-design-photoshoot-director; generation is as-video-ai-director.

Render a real MP4 right inside Cowork using Pillow + ffmpeg. No Remotion, no GPU, no network. For fast, DISPOSABLE social stings, quote cards, and kinetic-text shorts when the Remotion pipeline can't run (e.g. the Cowork sandbox).

Turn any product URL into a full animated demo/launch video (Remotion/Hyperframe), scrape branding + real product images, rebuild the core UI, simulate a demo, show a showcase + CTA.

Run a productized AI-video service as a business, attract clients with a 4-part "revenue" video (credibility hook → real proof → method reveal → DM call-to-action), close in the DMs with a paid consult (no free calls), deliver with a 6-step staged-approval SOP (idea → script → cheap character render → produce → 90% cut with music → final).

Generate a punchy 10-20s vertical short (Remotion/Hyperframe) from a hook, tip, stat, or quote, pattern-interrupt open, kinetic text, burned-in captions, loop-friendly close.

End-to-end "drop a raw file, get a finished video" studio (video-use (transcript-first editing) cuts filler words, silences, retakes on word boundaries) then a word-synced motion-graphics engine of your choice, HyperFrames (HTML-native, default) or Remotion (React). It asks which.

Turn an image + a script into an AI talking-head (avatar) video for Reels/Shorts/TikTok/ads, generate the voiceover (TTS), drive a talking-person render (HeyGen/D-ID-style), then caption, add B-roll/overlays, and format vertical. For a presenter-style video without filming.

Turn a business's real Google reviews into an animated testimonial/social-proof video (Remotion/Hyperframe), scrape rating + reviews, animate stars + a review carousel + proof counters.

Convert any raw spoken script into TTS-ready narration that sounds human on any AI voice engine (the own-stack voice lane (CoreReflex cloned + stock voices) first, vendor engines (ElevenLabs-class) as fallback only) via strategic ALL-CAPS emphasis, dramatic ellipsis pauses, staccato rhythm, question lifts, and authority periods, while preserving every claim, statistic.

Optimize content to get cited by AI answer engines (ChatGPT, Perplexity, Google AI Overviews, Claude), answer-shaped structure, FAQ/answer blocks, quotable sourced stats, entity clarity, schema, llms.txt, and a citation-share audit. The answer-engine extension of the SEO domain, pairs with as-seo-optimizer.

App Store Optimization for Apple App Store + Google Play, store keyword research, title/subtitle/keyword-field metadata, icon/screenshot/preview creative sets with in-store A/B tests, ratings-and-reviews velocity + reply, listing CRO, Apple Search Ads basics, and a monthly ASO report. The in-store algorithm + store-listing conversion discipline, distinct from web SEO and web ads.

Mine Google Search Console for CTR leaks and striking-distance queries (every query, not just questions) score missed clicks against a labeled expected-CTR curve, flag query×page collisions, and route title/FAQ/refresh work to the producing skill. Real GSC numbers only; does not invent rankings or promise traffic lifts.

Operate Google Search Console (submit + monitor sitemaps, request indexing (honestly), read search performance (queries/pages/positions/CTR + the striking-distance set), inspect URLs for real index/coverage status, and diagnose not-indexed reasons) via the Search Console API with the service-account auth pattern (Workspace blocks OAuth).

Audit every Google Search Console property the calling identity can see, flag unverified sites, tag lead-gen vs product, and only then run question-demand and CTR-gap on the readable ones. Real list_sites data only; never invents queries for a siteUnverifiedUser property.

Run the full organic-SEO growth system end to end, keyword research (filtered for winners), blog posts at scale for topical authority, money-keyword service pages, on-page + technical SEO, deploy, and Search Console indexing. Entry point for the SEO domain.

Reverse-engineer the businesses already winning the local map pack (live GBP competitor forensics (categories, attributes, review velocity, response + posting patterns, photos, services), website gaps (keyword gap, page-2 goldmine, service×city coverage), review sentiment mining, citation consistency, and entity buildout) into a prioritized, evidence-backed 90-day plan with velocity targets.

Audit and optimize an existing page or post for on-page SEO, titles, meta, headings, internal links, schema, snippet targeting.

Run keyword research, cluster planning, and content-gap analysis for Agency Script.

Build and publish a programmatic SEO site or page-set at scale, research keywords/clusters with real data (DataForSEO), derive each page's target word count + outline from the live SERP, generate original useful pages (incl. local/city expansion + affiliate "best of"), then publish to GitHub Pages/Vercel/WordPress, submit the sitemap, monitor indexing.

Mine Google Search Console for the questions people already type (the same Custom (regex) filter the Performance dashboard uses, plus cost, comparison, and problem lenses) then score each query as lead-gen demand and route it to FAQ/schema, a blog, a calculator, a comparison page, a lead magnet, Google Ads copy, or AEO. Real GSC numbers only; does not submit sitemaps or invent rankings.

One input (a domain/URL, optional competitor) → a complete, branded, client-ready SEO audit report (PDF/HTML), crawl + technical checks, keyword/gap vs a competitor, on-page scoring, and a prioritized fix list (quick wins vs strategic). The classic paid front-end offer, end to end.

Audit and fix the platform-level SEO foundation, rendering strategy (static/ISR/dynamic), caching, Core Web Vitals, crawl budget, sitemaps, and cache invalidation for the Next.js site.

Build industry/vertical landing pages and programmatic-SEO page sets (e.g. "AI for real estate", "selling AI to utilities") on-brand and conversion-ready.

Run a full Generative Engine Optimization audit on a site, score how well it gets cited by AI answer engines (ChatGPT, Perplexity, Google AI Overviews, Gemini) and return a prioritized, client-ready findings set. Orchestrates citability scoring, AI-crawler access, llms.txt, schema, platform readiness, and AI-visibility checks into one scored report input.

Measure a brand's visibility inside AI answer engines. Run the client's key buyer prompts through ChatGPT, Perplexity, Google AI Overviews, and Gemini and record whether the brand is mentioned, cited (linked), the sentiment, and which competitor gets named instead. Tracks share-of-AI-voice and the citation gap vs competitors over time.

Score how citable a web page is for AI answer engines (ChatGPT, Claude, Perplexity, Gemini) (a 0-100 citability score across answer-block quality, fact density, structure/extractability, and E-E-A-T) then return specific rewrites that make passages easy for AI to extract and quote. Built on the research that AI cites self-contained, ~130-170-word, fact-rich, answer-first passages.

Run the Generative Engine Optimization (GEO) service end to end, get a site cited by AI answer engines (ChatGPT, Perplexity, Google AI Overviews, Gemini, Claude) and sell it as a productized agency offering. Routes the GEO workflow: prospect → full GEO audit → citability + technical + schema + llms.txt fixes → platform-specific tuning → AI-visibility monitoring → client report + retainer.

Optimize a site for the specific AI answer engines that matter to the client (Google AI Overviews, ChatGPT/SearchGPT, Perplexity, and Gemini) because each cites differently.

Turn a GEO audit into a polished, client-facing Generative Engine Optimization report (Markdown + PDF), combine citability, technical/AI-readability, schema/llms.txt, per-engine readiness, and AI-visibility-vs-competitors into one branded deliverable with an overall score, business-impact narrative, and a prioritized fix roadmap (quick wins vs strategic).

Make a site machine-legible to AI answer engines, audit and fix AI-crawler access (robots.txt/meta for GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot), JS-rendering risk, the llms.txt standard, and Schema.org structured data (detect, validate, generate correct JSON-LD). You can't be cited if the AI can't read you.

Marketing front door and stage diagnostic. Marketing is not one domain, it is six stages across roughly 138 skills, and the expensive mistake is fixing the wrong one.

Build and optimize profitable Google Ads search campaigns end to end, Single Keyword Ad Groups (SKAG), keyword research, bulk ads with pinned headlines, negative keywords, matched landing pages, conversion tracking + remarketing, and a ROAS audit loop.

Plan and run a complete paid-acquisition program across every channel. Pick the channels that fit the offer and funnel stage, stand up the tracking spine first, then route to the right per-channel builder and govern the whole account by profit.

Package a rendered Meta ad batch into an Ads Manager bulk-import kit, one spreadsheet row per ad with vetted copy, UTM'd destination URL, resolvable image reference, and an explicit PAUSED status, plus the packaged PNGs and a README. The route that works when an ad account isn't API-writable (Ads MCP not enabled) or when the owner wants to review the whole batch before it touches the account.

Turn one offer doc into a batch of scroll-stopping, policy-safe Facebook/Instagram static ad images, each built from a proven hook formula (How-To Speed, Bold Claim, Pain Question, Humor, Curiosity), a battle-tested visual template (Superhero Punch, Old Way vs New Way, Red Hot Offer, Pain to Paradise, and 6 more), and a 4A buyer lens (Actionable, Analytical, Aspirational, Anthropological).

Deploy a rendered Meta ad batch straight into a Facebook/Instagram ad account through the Ads MCP API, preflight the account (API-writable, page attached, pixel present, audiences populated), upload images, build creatives, and create campaign/ad set/ads all PAUSED for owner review and manual activation. Reports blockers instead of substituting accounts.

Diagnose, build, and scale Pay-Per-Lead / Pay-Per-Call lead-gen campaigns, classify the campaign type, find the ONE binding constraint in the pipeline (vertical → buyer → offer → funnel → traffic → lead quality → delivery → buyer conversion → scale), fix the highest-leverage bottleneck, then scale.

A 24/7 customer-success support assistant for pay-per-lead / pay-per-call / client-acquisition / lead-generation customers, get them unstuck fast by classifying the request (client-acquisition vs lead-generation), diagnosing the bottleneck, giving a direct fix + ≤6 action steps, citing the right Coda or vendor (Heyflow/LeadsHook/HighLevel/Twilio) resource.

Inventory and rewrite range, kWh, mpg-e, safety, 'self-driving', and ROI claims before ads spend. General brand claims stay as-core-brand-voice-guard; medical verticals route to as-vert-medical-dental.

Design the warm-audience retargeting strategy that recaptures people who already raised a hand, define the audience ladder (site visitors, video engagers, cart/checkout abandoners, email list, lookalike seeds), sequence the messaging by funnel stage (reminder to objection to proof to offer), and run it across RLSA on search, Meta retargeting, and omnipresence, with frequency caps and exclusions.

On-demand, any-platform direct-response ad SCRIPT writer, usable, filmable, conversion-focused scripts with a full framework library (AIDA, PAS, Before-After-Bridge, Hook-Proof-Offer-CTA, Problem-Demo-Result-CTA, UGC Testimonial Arc), 10 hook types, timed 15s/30s/60s structures, two-column visuals tables, a humor module, and a claims/policy gate.

Join Google Search Console commercial queries to an optional Google Ads search-terms export, promote organic cost/mover demand into phrase SKAGs, add junk (distance, jobs, DIY, operator scrapes) as negatives, and flag overlap when paid data is present. Does not create campaigns or spend; Ads overlap is labeled NEEDS PROOF until a search-terms export is supplied.

Render approved ad angles into actual download-ready PNGs with a re-themeable HTML/CSS design system, 9 visual templates, an 8-token brand swap, and headless-Chrome rasterization at 1080x1350 and 1080x1080. The render layer between as-ads-meta-creative (what to say) and as-ads-meta-deploy-mcp (push to the account).

Build TikTok and Reels paid campaigns, native hook scripts that feel like organic content, Spark-ad angles (boost real creator posts), and a creative testing matrix mapping hooks to 4A buyer lenses (reusing the Meta-creative hook/lens + policy discipline). Scripts + shot lists handed to the video pipeline; copy passes the policy + claims gate.

Build the conversion-tracking spine for a paid funnel end to end, measurement plan (events to KPIs), GTM container wiring (dataLayer, allowlist trigger, GA4 event tag) so events actually reach GA4, Meta Pixel plus Conversions API (CAPI) with event_id dedup, GA4 plus Google Ads key events, gclid/fbclid to CRM to offline-conversion import.

Turn one offer into a batch of UGC-style short video ad scripts (creator-authentic, hook-led, with shot lists and on-screen hook/caption specs) handed to the video pipeline to render. Each ad pairs a proven hook formula with a 4A buyer lens (reusing the Meta-creative hook/lens discipline) and passes the platform-policy + claims gate before it ships.

Build YouTube video ad campaigns end to end, hook-in-5s scripts for skippable in-stream and 6s bumper formats, targeting, and the profit-first ROAS audit frame from the Google Ads playbook. Scripts + shot lists handed to the video pipeline; campaign copy passes the policy + claims gate.

Generate a complete, coherent launch package from ONE message spine, landing page, ads, creative, social posts, email sequence, and post-sale surfaces that all say the same thing in each channel's native shape.

Build an executable campaign brief (objective, audience, message, channels, calendar, budget, and measurement) and plan field-event motions across hosting, sponsoring/exhibiting, speaking, and attending. Webinar production routes to as-launch-webinar-producer.

Build a multi-day challenge funnel end to end, the ONE transformation, the reg page, a daily arc of quick wins + cliffhangers, the engagement loops (community, email + SMS reminders), the offer-day bridge from the free or paid ($7-$97) challenge to the paid offer.

One-pass seeding of a product launch across ~55-70 named vetted launch, AI-tool, and startup directories (plus a Tier-4 long-tail to batch), an asset kit, per-site submission specs, and an acceptance tracker that earns referral traffic, a diverse backlink spread, and an LLM-citation surface (a GEO and SEO play, not just clicks).

Turn a validated live launch into an always-on, automated evergreen funnel that sells every day, a just-in-time (JIT) automated webinar, a REAL personalized perpetual deadline enforced at the link level (cart-close / bonus expiry / price step that actually changes for that lead), and a per-lead evergreen email sequence triggered from each lead's own entry date.

Plan and run a multi-week launch (founding cohort, product, or feature) with phased timeline, channel plan, content, and metrics.

One input → a finished sales webinar or VSL. Paste an offer URL (or describe the offer in a few sentences) and get the complete package in one session (extraction, angle slate, full script on the proven house frameworks, generated slide deck with speaker notes, and TTS-ready narration on the own-stack voice lane) minutes instead of weeks, no copywriter, no camera, no agency.

Plan and run an earned-media PR campaign across three phases, pre-launch (SMART goals + KPIs, audiences, owners + budget + timeline, finalize story/channels/assets, build the journalist list), launch (run-of-day schedule, monitoring + social listening, real-time optimization, engagement), and post-launch (KPIs beyond coverage to traffic/leads/sales, stakeholder report, learnings).

Produce a webinar end to end, promo, registration page, reminder + follow-up emails, slide script, and run-of-show.

Consult a CEO: bring a company-level decision and get an accountable answer, not a menu.

Consult a CFO: bring a money decision and get an accountable answer, not a spreadsheet.

Consult a CMO: bring a demand problem and get an accountable answer, not a channel list.

Consult a CTO: bring a technical decision and get an accountable answer, not a survey.

Consult a DevOps/SRE: bring a reliability, release, or infrastructure-cost decision and get an accountable answer.

Consult the executive team: bring a decision and get the right lens on it, or convene a panel when it is genuinely cross-functional. Routes to the CEO (is this worth doing), CTO (will it work and stay changeable), CMO (where demand comes from), CFO (can we afford it and do the economics work), DevOps (will it stay up) or QA (will it break what works).

Consult a QA lead: bring a release, a suite, or a scare and get an evidence-based verdict, not reassurance.

Apply Ive-era Apple house language (object on infinite white, unibody restraint, material honesty, huge margin) to a product brief, page, or shot. Not the Apple logo.

Apply Jeff Bezos's public method (work backwards from the customer, write the press release and FAQ before the build, Day 1, narrative memos) to a feature, page, or spec brief. Not a likeness, not "I am Jeff." Use for "Bezos thought process", "working backwards", "PR/FAQ first", "Day 1 memo", "six-pager", "empty chair customer".

Apply Rams-era Braun house language (mute functional objects, controls where the hand expects them, pale field, one quiet accent) to a product brief, shot, or UI. Not the Braun wordmark.

Apply Disney house language (worlds with a threshold, a weenie, theatrical presentation, backstage stays hidden) to a film, park-like experience, or page. Not the mouse, not the castle mark.

Apply Gary Vee's public content method (document don't create, jab-jab-jab-right-hook, spoken cadence, platform-native volume) to a post, cadence, or script. Not a likeness and not "I'm Gary." Use for "Gary Vee thought process", "Gary Vaynerchuk", "document don't create", "jab jab jab", "day-trade attention".

Apply Alex Hormozi's public offer method (value equation, grand-slam stack, proof before promise, specific numbers) to a page, VSL, or offer brief. Matches "Hermozi" too.

Front door for iconic figure and house overlays. Figures: Jobs, Hormozi, Gary Vee, Musk, Bezos, Rams, Ogilvy, Walt Disney, Thiel.

Apply Elon Musk's public thought overlay (physics sentence, engine vs fairing, delete before automate) then hand the teardown to the house algorithm skills. Not a likeness, not "I am Elon", not classified internals.

Apply 1960s-70s NASA house language (civic mission modernism, constraint as romance, hardware of a nation, diagram-as-aesthetic) to a film, graphic, or mission brief. Not the NASA seal or worm mark.

Apply Nike house language (athlete as proof, motion crop, one imperative line, the product is the verb) to a campaign, spot, or hero. Not the swoosh.

Apply David Ogilvy's public method (research first, the headline is 80% of the spend, sell don't entertain, specifics out-pull generalities, long copy only when every line earns the next) to an ad, page, or letter. Not a likeness.

Apply Pixar house language (one emotional question, want vs need, specific not generic cute, comedy from character) to a script, film brief, or story page. Not the lamp mark.

Apply Porsche house language (form is the function remaining, no substitute, heritage machine, analog honesty) to a product hero, film, or page. Not the crest.

Apply Dieter Rams's public method (less but better, the ten principles as vetoes on the object) to a product, UI, or brief. Not a likeness and not Ive-era Apple white.

Apply SpaceX house language (hardware-rich, stainless, high-bay, night, flame, test-fail-fix as the story) to a film, industrial brief, or shot list. Not the SpaceX wordmark and not the five-step algorithm.

Apply Steve Jobs's public method (subtract until one idea remains, the demo is the argument, concrete nouns over adjectives) to a brief, keynote, product page, or script. Not a likeness, not "I am Steve Jobs." Use for "Steve Jobs thought process", "Jobsian", "one more thing structure", "think like Jobs", "simplify this like Jobs".

Apply Stripe house language (docs as product, economic infrastructure said quietly, precise web craft, code as the demo) to a site, docs page, or product narrative. Not the Stripe mark.

Apply Tesla house language (product as the demo, Highland calmer tighter fewer parts, stainless night, one accent) to a product brief, UI, or shot. Not the Tesla wordmark and not a PPL site build.

Apply Peter Thiel's public method (0 to 1, a secret few people agree with, monopoly not competition, definite optimism as a plan) to a positioning, GTM, or strategy brief. Not a likeness.

Apply Walt Disney's public method (plus-it every beat, the storyboard is how you think, weenies pull attention through a path) to a script, ride-like experience, or film brief. Not a likeness, not the mouse.

Protect against data loss: back up the database, R2 assets, and code/config, set RPO/RTO targets, and run restore drills so recovery actually works.

Set up the CI quality gates that block bad code from reaching production, lint, typecheck, tests, and build in GitHub Actions, with branch protection, so only green code reaches the git-push-to-hostssh deploy.

Operate Cloudflare as the DNS + R2 object-storage + edge layer fronting a self-hosted fleet, DNS via a scoped Zone:DNS:Edit token + REST (the MCP has no DNS tools; touch only apex A + www, never MX/TXT/DKIM/SPF), R2 buckets via MCP + objects via the S3 API (the zero-egress restic backup target + per-app media buckets), proxy/TLS (SSL Full strict + Origin certs), and Workers/KV/D1/Hyperdrive.

Operate Coolify, the popular open-source self-hosted PaaS, for agencies and clients who run it, onboard a box, deploy apps from git (Nixpacks/Dockerfile) with automatic Let's Encrypt, give a service a custom domain (the FQDN gotcha), stand up one-click databases, wire backups, and cut DNS over.

Govern cloud + AI spend across the whole stack (the VPS fleet, R2, GPU/render, Claude API, and third-party adapters) with attribution, budgets, alerts, and unit-economics (cost per render/member).

Plan and apply Prisma/Postgres schema changes safely, classify additive vs destructive, back up, dry-run, apply with migrate deploy, and keep a rollback path.

Ship a release safely on the Next.js/hostssh/Prisma stack, pre-deploy checklist, git-push-to-hostssh → blue-green deploy → prod, migration sequencing, env sync, post-deploy verification, and rollback.

Manage environment variables and secrets across local, the app's hostssh container env, and the GPU box, audit parity, rotate keys safely, and keep secrets out of the repo.

Operate Google Cloud via the gcloud CLI + gcloud MCP, and set up the service-account auth pattern that makes Google-product MCPs (GA4, Search Console, Vertex) work when a Workspace org blocks unverified OAuth, create a least-privilege service account + key (chmod 600), grant roles, enable APIs, mint a scoped token for Admin-API writes, and ALWAYS restore the human account after activating an SA.

Operate Hostinger (the VPS provider our fleet runs on, plus managed WordPress/shared hosting, domains, DNS, and email) through the hostinger-api MCP. Provision and recreate VPS boxes, manage the cloud firewall and snapshots, buy and point domains, edit DNS safely (without breaking email), and drive the periodic renewal migration.

Run any DevOps task end to end for the Agency Script stack, confirm the production-change gate, then route to deploy, migrations, infra, observability, incidents, CI, secrets, backups, or cost. Entry point for the DevOps domain.

Run a production incident end to end, assess severity, mitigate/rollback, communicate, and write a blameless postmortem.

Set up and run monitoring for the Agency Script stack, Sentry errors, logs, uptime, render-queue health, and AI-budget/cost alerts with sane thresholds.

Operate the video render + media-generation queue, dispatch jobs, monitor progress, retry failures, scale GPU workers, and control per-render cost.

Provision and operate the self-hosted GPU engine fleet (Voicebox, ComfyUI, whisper.cpp, MuseTalk, Remotion render workers) as services behind the MediaProvider seam, reachable from the app via Cloudflare Tunnel + Access.

Pick, license-check, and compose a short own-the-stack (analytics, wiki, helpdesk, email, LMS, search, identity) into a hostssh deploy spec. Not the GPU box, not VPS capture/restore, not Coolify, not Paperless-only, not make-vs-buy.

Operate a self-hosted app fleet on hostssh (our own server control plane) instead of managed hosting (Vercel) and managed Postgres (Neon). Deploy apps (git-push or dashboard), stand up managed databases on a private mesh, capture whole-server encrypted.hsi images and restore/clone/relocate a box in one step, run restore drills, cut DNS over revenue-site-last, and run the periodic VPS renewal.

Audit our OWN agent surface for security drift, over-broad autonomous:true on destructive/outward skills, missing G7/G8 human gates, dangling canon or skill refs, secrets committed in skill files, and role/automation/name mismatches. Read-only; never edits configs.

Runtime BEHAVIORAL anomaly detection for an agent that was allowed-but-is-now-acting-wrong, hijacked mid-session by injected tool output or poisoned context. Baseline normal, arm tripwires (scope/tool shift, new-host egress, secret access, off-task loop), confirm with a human before the sensitive action, then quarantine or kill it.

Build the ephemeral isolation boundary an untrusted agent run executes inside (a disposable non-root container, no network after build, read-only host mounts, secrets injected at runtime not as CLI flags, one writable output volume, destroy-after-run) plus a provider matrix (Docker, own VPS, cloud VM, edge unsuitable).

Test your own API and auth layer against the OWASP API Top 10, broken object/function-level authorization, JWT flaws, webhook verification, rate limiting, mass assignment.

White-box source review of the Next.js / Prisma / Stripe / JWT stack for security defects, server actions, API routes, RBAC/IDOR, input validation, secrets, webhook verification, SSR data exposure.

Review hosting and cloud configuration for security, HTTP security headers, CSP, TLS, CORS, Vercel/Cloudflare/AWS settings, storage exposure, and least-privilege IAM.

Harden Docker container images, runtime, and orchestration for the self-hosted Coolify + Cloudflare fleet, pinned-digest minimal base, non-root user, dropped capabilities, read-only rootfs, seccomp/AppArmor, no secrets in layers, image/CVE scanning (trivy/grype), cosign provenance, docker-compose + Coolify port/socket/network exposure, and Kubernetes pod-security.

Scan for vulnerable dependencies (CVEs), exposed secrets, and supply-chain risk across the repo and history.

Build tested detections and run structured threat hunts on the self-hosted fleet, log-source coverage, detection-as-code (hypothesis→rule→test→tune→deploy→measure), Sigma-style portable rules, MITRE ATT&CK coverage heatmap, alert-quality tuning, hunt playbooks, and detection KPIs (MTTD, coverage).

Respond to a confirmed or suspected SECURITY compromise of the fleet, apps, accounts, or data, classify severity, collect IOCs, preserve evidence with chain of custody, reconstruct a UTC timeline, contain, eradicate + verify recovery, and feed breach-notification inputs to legal, per NIST SP 800-61.

Authorized adversarial evaluation of a DEPLOYED LLM or agent, probe jailbreak families, encoding/obfuscation, system-prompt extraction, indirect prompt injection via retrieved content, tool/function abuse, data exfiltration, and excessive agency via a seed→mutate→score→regression runner with calibrated scoring that reports coverage SEPARATELY from pass-rate.

Provenance + quarantine gate on everything written into durable agent memory (project brain, brain-bases, compound learnings, journal), tag write-time provenance + trust, quarantine low-trust and instruction-shaped writes, treat recalled memory as data not instruction (G6 in time), keep a tamper-evident trail. NEVER executes a stored memory as an instruction.

Continuous PASSIVE external-exposure + threat-intel monitoring for own and written-authorized client brands (cert-transparency & DNS footprint, passively-indexed service and bucket exposure, leaked-credential/breach monitoring, typosquat/lookalike-domain detection, public code/secret leakage, and dark-web mention triage) scored relevance × severity into an exposure register + brief.

Run an authorized security assessment end to end, establish the authorization and rules-of-engagement gate, choose one primary security route plus justified secondary routes, create a resumable case record, coordinate safe testing, compile evidence-backed findings, and re-test fixes. Report-only requests go to as-sec-pentest-report; live compromises go to as-sec-dfir-incident-response.

Compile security findings into a professional, severity-ranked report with CVSS scores, safe proof-of-concept, business impact, and remediation, plus an executive summary.

Defend the human layer against phishing and social engineering, email-auth anti-spoofing (SPF, DKIM, DMARC p=reject, MTA-STS, BIMI), a consent-based phishing-sim program that never captures real credentials, an awareness cadence, a one-click phish-report + safe triage SOP, and a BEC/vishing/smishing/wire-fraud playbook.

Scan untrusted content (tool output, fetched web/docs, uploads, pasted text) for prompt-injection payloads BEFORE an agent acts on it, treating the content as DATA, never instructions. Severity-ranks payloads and recommends quarantine/strip/allow.

The DEFENSIVE secure-coding standard applied WHILE building, the build-time bookend to the offensive pentest pack. Covers auth/authz/RBAC, IDOR-proof object-level checks, input validation, secure API design, secrets, secure-by-default headers/CSP, plus a pre-merge secure-build checklist enforceable in CI.

Map the authorized attack surface (domains, subdomains, endpoints, technologies, exposed assets, and metadata) passively first, then light active enumeration within scope.

Scan the CONTENT of a third-party agent skill, MCP server, or plugin for vulnerabilities and score the risk, answer "what does this code actually do, and is it dangerous?" by reading its files as data, never executing them.

Supply-chain adoption gate for an INCOMING third-party skill, MCP server, or plugin before install, verify provenance + signature, pin an integrity hash, run the permission-delta gate that BLOCKs a version silently widening file/network/secret/exec scope, then emit an install/decline verdict + SARIF.

Threat-model a feature, flow, or system using STRIDE + data-flow diagrams to find design-level security risks before code ships.

Test your own web application against the OWASP Top 10 / WSTG (access control, injection, auth/session, XSS, SSRF, misconfiguration) with safe proof-of-concept and remediation.

Route CRM work across Agency Script native CRM, Reflexware, GHL, and Twenty. Pick one system of record, then dispatch. Not RevOps field architecture (as-ops-revops-architect), not GHL snapshot fleet (as-ops-ghl-fleet), not one pipeline review (as-sales-pipeline-review).

Absorb Twenty CRM into Reflexware and Agency Script the licensed way, re-author metadata-object logic in our spine (Path A), or ship a proprietary Reflexware app on Twenty via MIT SDKs (Path B), or run unmodified Twenty as a separate AGPL service (Path C). Never copy twenty-server or twenty-front into our app; never take Enterprise-marked files.

Classical responsible-AI and algorithmic-fairness review, audit a model or automated decision system for bias and fairness (demographic parity, equalized odds, disparate impact), write model cards, run an algorithmic impact assessment, and classify it under EU AI Act risk tiers with NIST AI RMF and ISO 42001 control-ID mapping.

Generate and maintain robust developer and user documentation for the app, populate /docs/dev and /docs/user from the codebase, planning docs, and commit history, and power the in-app docs/support section.

Deep-dive production-readiness audit of the whole app (pages, API routes, UI/UX, security, schema, PRD gaps, placeholders vs real functionality) scored on the HOUSE instrument, so every category normalises 0-100, a weighted composite, and an honest denominator where an unassessed factor is reported as a gap instead of scored zero.

Run the board meeting machine and keep the governance record clean, deck construction on the narrative arc (performance vs plan → the one decision → asks), pre-reads sent early so meetings decide, not present, annual governance calendar (approvals, comp, audit, budget), minutes and resolutions drafted to counsel-review standard, action tracking, committees, and director onboarding.

Generate, consolidate, and maintain the Bootspring build queue, turn preseed + planning docs into thousands of unchecked Bootspring-format tasks in planning/TODO.md, merge task_queue.md into a single source of truth, and keep the build loop fed.

The step-0 foundation that makes a new venture a legitimate, payment-ready business, a plain-English orientation + checklist for setting up a legal entity (LLC/etc.), an EIN, a business bank account, and payment processing (Stripe/GHL), with a vendor-neutral comparison of formation paths, a business-naming strategy, the order of operations, and the ongoing-compliance budget.

Build a robust, long-format, INDUSTRY-SPECIFIC business plan via a multi-step full-context Q&A, uncover the business one topic at a time (industry first), assemble a tailored long-format outline the owner approves.

Analyze workload vs available capacity and forecast utilization, then recommend automate / deprioritize / hire. Distinct from as-ops-status-report (project status). This is the bandwidth math.

Mine a screenshot library (Snagit, CleanShot, macOS screenshots) for proof by reading each capture's embedded OCR metadata first, so thousands of images become greppable text instead of a manual scroll. Then triage, redact sensitive strings, and publish.

Write a change request with impact analysis, rollback plan, CAB-style approval, and comms. Binds the production-change gate when the change touches the app/infra (owner runs it).

Cut CI/CD and hosting spend (relax or gate GitHub Actions workflows, stop redundant runs, disable costly crons, and tame Vercel auto-deploys) without breaking the safety nets that matter.

Track competitors and produce battlecards and positioning gaps.

Take one competitor name/URL and produce a polished, client-facing competitive teardown, scraped positioning/pricing/offers, a feature + messaging matrix, gap + threat analysis, and a presentation-ready render that maps the rival to the assurance gap they leave. Not an internal battlecard: a leave-behind a prospect can read.

Track the compliance program and audit readiness, requirement inventory, control owners, evidence, and gaps for SOC 2, ISO 27001, and GDPR/CCPA. Distinct from as-ops-legal-compliance (reviews a specific contract/policy) and as-sec-* (tests controls). This is the compliance-program tracker.

The comms layer when something goes wrong (outage, data incident, viral complaint, bad press, layoffs) severity, activation, holding statements (what we know / what we don't / next update), stakeholder sequencing (affected before press, staff before public), drafts per channel (status page, email, social, press, internal), cadence contract, one-voice spokesperson law, do-not-say list.

Durable Architecture/Operating Decision Records (ADR/ODR) (the WHY behind a costly-to-reverse choice) context, options considered, the decision, consequences, reversibility (Type-1/Type-2), who decided, and what was rejected, mined from meetings and sessions into an append-only log with a proposed/accepted/superseded lifecycle. The institutional memory an agency loses when contractors churn.

Stand up and run a self-hosted document-management archive (Paperless-NGX-class), the tag / correspondent / document-type taxonomy, the OCR → auto-classify → tag → full-text-index pipeline, faceted retrieval, and Restricted-class governance (least-privilege, audit-logged, human-gated retention) on owned hardware.

Run and grow an online store (Shopify/DTC/marketplace), model the unit economics (AOV, COGS, margin, CAC, LTV, contribution margin), write converting product listings + merchandising, manage catalog/inventory + reorder points, set fulfillment + a lawful returns flow, run store-funnel CRO + abandoned-cart recovery, and pair acquisition with an email/SMS retention engine (where DTC profit lives).

The end-of-day report you can actually send, to an employer, a client, or a stakeholder. Turns the day's real work into a short, evidenced, client-safe update covering what was accomplished with proof, what moved, what is blocked and on whom, and what happens tomorrow.

Stand up an EOS-style operating system for an agency end to end, the two-page Vision/Traction document, an accountability chart (seats and roles before names), a weekly scorecard of 5-15 measurables, quarterly Rocks with owners, a single prioritized issues list, and the Level 10 weekly meeting agenda that runs them all on cadence.

One input (a live site URL + optional repo) → a measured, calibrated 1-100 experience scorecard for the whole web property, marketing site and app. Runs a real tooling harness (Playwright, axe-core, Lighthouse, Core Web Vitals, security-header probes) across desktop/mobile, scores 17 dimensions, calibrates against a best-in-class bar, and renders a branded scorecard + prioritized roadmap.

Run revenue and billing operations, MRR/revenue tracking, failed-payment recovery, refunds, basic reconciliation, and finance reporting.

Question every requirement against physics, rewrite the spec so only the engine, surviving regulation, and owner-confirmed taste remain. Past-defect RCA is as-ops-root-cause; order of work is as-ops-the-algorithm; PRD writing is as-pm-write-spec.

Diagnose and run a business as four sequential pillars (Value Creation, Marketing, Sales, Value Delivery) find the ONE binding constraint and pull that pillar. Leaves EOS/L10 to as-ops-eos-operating-system, moat/strategy to as-ops-strategic-advisor, and past-defect RCA to as-ops-root-cause.

Package and run fractional CMO/COO/CFO/CTO services as a productized agency line, role-scope matrices (owns vs advises vs never touches), retainer tier math, the 90-day arc (diagnose → quick wins → operating rhythm), per-role artifacts + client scorecard, capacity math, and hire-your-replacement transitions inside counsel-gated liability boundaries.

Prepare and run a capital raise + ongoing investor relations, decide whether/how much to raise and tie use-of-funds to milestones, build honest materials (deck, assumption-backed model, one-pager, a controlled data room), target + track investors through a tight process, get cap-table literacy, and run a candid post-raise investor-update cadence.

Point it at anything (an app, a website, a repository, a document set, a process) and get a complete gap analysis covering what exists, what is missing, what it costs to leave open, and who owns closing it. Domain-agnostic where the specialist audits are domain-locked, and it routes into them when depth is warranted.

Operate a fleet of GoHighLevel sub-accounts via the v2 REST API and productize an agency system into a sellable GHL snapshot.

Extract one shared booster and many vertical fairings from a fleet of duplicate parts. PPL capture extract stays as-ops-lead-gen-tesla-build; this is the general platform move.

Run a 90-day goal sprint that turns an ambitious target into a daily activity number you actually hit, set the quarterly goal, reverse-engineer it to the daily activity from real conversion math (e.g. revenue goal → deals → consults/day at your close rate), make it public with an accountability partner, track the daily number, review weekly, then reset and repeat.

Run the compliance-heavy grant and RFP/RFI/RFQ response machine, bid/no-bid scoring with the courage to no-bid, the compliance matrix before any writing (non-compliance is the #1 disqualifier), fresh response-library reuse, outlines mirrored to evaluator weighting, pink/red/gold reviews, grant eligibility/budget/reporting discipline, receipted submission ops.

Run a general hiring and recruiting system for one role end to end, a role scorecard (mission, outcomes, competencies) before the job post, an on-brand job post and sourcing plan, a screening rubric and funnel, a structured interview scorecard, a paid test task that proves the work, and an offer/decision package.

Operations category hub: routes to the right leaf for running the company (runbook, process, status, decisions, meetings, goals), money and governance (finance, fundraising, board, procurement, vendors, insurance, compliance, risk), people (hiring, people-ops, succession), hard-numbers strategy (first-principles, the-algorithm, make-vs-buy, idiot index, throughput).

Stand up one SSO/IdP in front of every self-hosted admin UI (Authentik default; Authelia or Keycloak as swaps) with OIDC, least-privilege apps, and Tunnel/Access in front. Not Cloudflare Access alone (as-devops-cloudflare), not password vault ops (as-devops-env-secrets-ops).

Compute should-cost vs actual (idiot index) for a part, process, SaaS seat, or deliverable and route delete / redesign / in-house / keep. Vendor TCO is as-ops-vendor-review; the buy process is as-ops-procurement.

Prepare the business-insurance coverage review a licensed broker acts on, plain-language taxonomy (GL, E&O, cyber, workers comp, property, D&O, EPLI, umbrella, media), exposure mapping from the business model, declarations extraction (limits, deductibles, exclusions, endorsements), gap/overlap flags, COI + additional-insured contract compliance, renewal package, claims readiness.

Run investor relations after the money is in (the monthly/quarterly investor-update system (locked metric set, lowlights before wins, bad news first, a cadence that never slips), data-room hygiene with a currency SLA and access log, an investor CRM (who invested what/when, follow-on capacity, intro graph), specific trackable asks, and information-rights/annual-meeting readiness) every figure.

Design a test article, abort criteria, telemetry, and one-variable next flight instead of a waterfall campaign. Full campaign plans stay as-launch-campaign-planner.

Turn raw conversations, chats, notes, and decisions into structured, reusable knowledge filed where the team can find it, a knowledge base in Notion (or the Project Brain / files).

Build and upgrade pay-per-lead / local-service lead-gen sites to SpaceX first-principles reasoning and Tesla Highland build quality, shared capture booster, vertical-specific fairing, panel-gap UI on the **site vehicle**, fail-closed lead API, CAPI, TCPA, prompt-before-delete. Owns the vehicle (the live site + capture path).

Review contracts and policies, run compliance checks, and route documents for e-signature for Agency Script.

One input (a business name + location) → a complete, branded, client-ready local SEO / Google Business Profile optimization pack, GBP optimization checklist, NAP/citation gap scan, category + local-keyword recommendations, review-response templates, and ready-to-publish local landing copy (via as-seo-vertical-page-builder). The "rank in the map pack" deliverable, end to end.

Classify parts as engine vs mass and verdict make / buy / delete. Vendor scoring is as-ops-vendor-review.

Make meetings produce outcomes, not just notes, prep a briefing before (agenda, context pulled from the knowledge base + prior notes + relevant research, the decisions the meeting must reach), then after, turn the transcript/notes into a clean recap with decisions, owner-assigned action items + due dates, and open questions, filed to the knowledge base and tracked to done.

Specify a live ops console: leading indicators, abort criteria, owners, daily/weekly go-no-go. Exec dashboards stay as-analytics-exec-dashboard; checklists stay as-ops-runbook.

Operate many clients' OWN/AUTHORIZED ad and social accounts without false cross-linking, authorization + ToS decision tree first, the platform's sanctioned multi-account access preferred over anti-detect; fingerprint-isolated profiles + per-client proxies + vaulted credentials as fallback.

Turn one plain-English automation request into a validated, importable n8n workflow JSON. Pick the platform (n8n by default when self-hostable), start templates-first, assemble the node graph with correct expressions, validate structure + expressions + connections + AI-agent nodes, and emit a deployable workflow.

Grow by node placement, occupancy, and uptime. Not more ads. Generic flywheels stay as-funnel-growth-loops.

Design the post-launch telemetry → patch → rollback loop for a live product or membership. Engineering compound loop is as-eng-compound-loop; pre-scale flights are as-ops-iterative-test-article.

Run the ongoing people/HR function after the hire, performance reviews on role scorecards, market-anchored compensation bands + raise/promotion criteria, a plain-English employee/contractor handbook + policies, time-off/leave tracking, performance-improvement plans, and a clean, kind, secure offboarding.

Design and build high-end, animated marketing pages and dashboard UI for the app (homepage, marketing site, and dashboard) with GSAP, Framer Motion, Aceternity, Radix, shadcn/ui, Tailwind, and motion-primitives, kept consistent entry-to-exit. Boundary: PPL/local-service lead-gen site vehicles (SpaceX algorithm, Tesla panel-gap quality, capture/CAPI) go to as-ops-lead-gen-tesla-build.

One input (a process described in prose) → a complete, hands-off process kit (extracted steps, a Mermaid flowchart with exception branches, a one-Accountable-per-step RACI, an edge-case register, and a numbered SOP a fresh operator runs with zero gaps) assembled into a single client-ready file. Distinct from as-ops-runbook (which WALKS a recurring cadence; this WRITES the procedure once).

Analyze a process and make it faster/leaner, map current state, find the bottleneck, eliminate or automate steps, quantify the win. Distinct from as-ops-process-doc (which documents). This one improves.

Run procurement as an operation, intake to PO, request intake + threshold triage (p-card vs PO vs RFP), sourcing to shortlist, RFP/RFQ construction with the weighted evaluation matrix frozen BEFORE proposals arrive, bid scoring on TCO not sticker price, negotiation prep (BATNA, levers, term targets), approval-ladder + PO design, and the renewal/SaaS-sprawl audit that kills maverick spend.

Agency Script OS orchestrator for taking a large product feature from idea to shipped, end to end (spec, system design, threat model, build, and launch audit) landing the spec in /planning.

Diagnose throughput: takt vs actual, WIP, bottleneck, changeover, a 14-day rate plan. Capacity hours stay as-ops-capacity-plan; which pillar to pull stays as-ops-four-pillars.

The State of the Union for a project, a complete, evidenced picture of where everything actually stands right now, scored on the house instrument. Inventories what is built versus planned, what is proven versus asserted, what changed since the last pass, and what is blocked, then ships it as a branded report with an honest denominator.

Run recruiting as a standing multi-role operation, not one-off job posts, headcount intake from the budget/FP&A, channel mix per role family, outbound sourcing sequences, referral mechanics, employer brand, a talent CRM with silver-medalist recycling, funnel instrumentation with leak diagnosis, interviewer calibration, agency management, and candidate-experience SLAs.

Architect the revenue system between marketing, sales, and success, lifecycle-stage models with entry/exit criteria and handoff SLAs, CRM field/pipeline architecture plus the field-sprawl audit, routing design, funnel-math instrumentation, forecast-hygiene stage definitions, stack rationalization, and the revenue-leak audit that finds where handoffs drop and data dies.

Identify, score, and mitigate operational/business risks. Build the risk register (severity × likelihood, mitigation, residual). Distinct from as-sec-threat-model (security/STRIDE) and legal-risk (legal). This is operational/business risk.

Structured root-cause analysis of a KNOWN, already-happened problem, rank defects by Pareto (80/20), drive 5-Whys and a fishbone/Ishikawa diagram to a single evidence-backed root cause, split contributing factors from the true cause, then write corrective + preventive actions (CAPA) closed by a verification-of-fix step. Numbers-forward.

Run the recurring daily/weekly/monthly/quarterly operations checklists and surface what needs the founder.

Numbers-forward simulation for business decisions (scenario architecture (baseline to disruption), decision trees with expected value, Monte Carlo over input distributions with confidence intervals, market-response curves, and constraint / system-dynamics models) each fronted by an honesty gate that states an accuracy tier, runs a named-bias pass (optimism, anchoring, base-rate neglect).

Run the operator-owned headed sitting that maps authenticated private HTTP (GHL, LinkedCamp, other NetJive targets) into a registry callable via API, MCP, or CLI. Not PyroSync Extractor Studio (living tables).

Turn a discussion transcript or feature brief into a passing SPEC.md, the gate a build must clear before code. One input → goals/non-goals, a numbered decision ledger (D-01..) with rationale and rejected paths, atomic testable requirements (R-01..), a requirement→test verification matrix, and the spec landed in /planning.

Produce a project/ops status report, green/yellow/red headline, KPIs vs target, progress narrative from the tracker, risks, and owned action items. Distinct from as-analytics-weekly-metrics (the business KPI scorecard) and as-design-internal-comms (the format library). This is a project/initiative status.

Brutally honest strategic analysis of the business/product, weaknesses and failure modes, the 20% of actions driving 80% of results, low-competition/high-demand opportunities, what's truly unique and how to 10x the moat, and billion-dollar expansion paths.

Survive the loss of any one key person, bus-factor inventory of who uniquely holds relationships, credentials, knowledge, and authorities; continuity runbooks per critical function (who acts, with what access, within what limits); break-glass design with dual control and audit; knowledge-transfer triage and operator manuals; interim vs permanent succession.

Inventory science and demand planning for ecommerce and manufacturing clients, ABC-XYZ segmentation, safety-stock and reorder-point review, OTIF (on-time-in-full) delivery analysis, and forecast-accuracy review (MAPE, WAPE, bias), where every recommendation states its assumptions and a service-level target instead of a false-precision number.

Design a complete team-member onboarding and training program for one role, a role ramp map, the SOP/training curriculum pulled from existing process docs, a structured 30/60/90-day plan with milestones, competency checks at each gate, an access/tools checklist, and the manager cadence that runs it.

Run Elon's production algorithm on any system, make requirements less dumb, delete (gated), simplify, accelerate, automate last. PPL sites stay as-ops-lead-gen-tesla-build; which business function to pull stays as-ops-four-pillars; past defects stay as-ops-root-cause.

Spec owned uptime checks and an optional public status page for the fleet, Uptime Kuma-class probes, Healthchecks dead-man for crons, ntfy/Gotify for pages. Not incident command (as-devops-incident-runbook), not Sentry triage (as-devops-observability-monitor).

Evaluate a vendor or tool on cost, risk, and TCO, then recommend renew / replace / renegotiate. Distinct from legal:vendor-check (the agreement-status/legal check). This is the business cost/risk evaluation.

Vendor a third-party tool or open-source repo into the app and fully white-label it, strip external licensing/branding, remove calls to third-party servers, rebrand, and make it a unique, dependency-free part of the platform.

Design and run the certification's adaptive proctored exam as a defensible measurement instrument (IRT-calibrated item bank, max-information adaptive selection, theta (ability) scoring with standard-error stopping, exposure control + content balancing, and aberrance analytics that flag for a human) making a static quiz employer-trustable.

Produce certification course content: module outlines, lesson scripts, labs, supplements, and assessment items.

Manage the credential lifecycle: issuance, public verification, revalidation cadence, expiry reminders, and revocation.

Assist human reviewers grading certification submissions against calibrated rubrics, pre-score, surface evidence, flag integrity issues, draft feedback.

Help a member or the team run one of the 5 core Scripts (Discovery, Architecture, Delivery, Optimization, Scale) and produce its deliverable.

Run WCAG 2.2 AA accessibility audits + the compliance program on any web property (ours or a client's) POUR-structured, layering screen-reader and keyboard-only passes over automated scanners (floor, not ceiling), severity tiering, remediation by user impact, VPAT/ACR prep, legal orientation (ADA, Section 508, EAA) without legal advice.

Install a deny-by-default RUNTIME firewall in front of an autonomous agent's tool calls plus a tamper-evident hash-chained ledger, then emit a signed attestation, agent X ran under constraints Y, touched Z, spent $W, verifiable against the chain hash. Adds a kill-switch to quarantine a hijacked agent mid-run.

Run a swarm of agents off a kanban board instead of a flat task list, triage each card into a runnable spec, let workers pull cards through Ready → Running → Review → Done under WIP limits, and (the part that actually matters) recover the cards that land in Blocked with a named escalation playbook so the swarm never silently jams.

Stop one-off prompting: design autonomous loops that do the work and supervise at a higher level. Picks the lightest mode on the autonomy spectrum (one-shot → bounded goal+verifier → interval loop → recurring automation → continual-learning), sets a stop condition + budget cap, gives the loop a way to verify its own work (the #1 reliability lever), and keeps a human on every irreversible action.

Design and run a standing autonomous AI organization, a business unit staffed by agents in named roles, not one prompt at a time.

Author the allow / ask / deny permission policy that defines an autonomous agent's autonomy envelope, the static guardrail every loop, org, and coding harness runs inside (e.g. a settings.json permissions block).

Run responsive + accessibility (WCAG 2.1 AA) gates on the Next.js app, breakpoints, keyboard nav, focus management on Radix/shadcn, reduced-motion, contrast, screen-reader semantics.

Score a named workflow on an L0-L5 autonomy ladder (manual → fail-closed autonomous) and write the one-level climb. Not vehicle Autopilot advice.

Turn a natural-language question about the Project Brain into a portable, diffable saved query/view (Obsidian Bases-style YAML) over the Brain's frontmatter-tagged markdown notes. Then render the current result and store the.base so it is rerunnable forever. Pairs with as-core-project-brain (Mode C) and the Retrieval Protocol; loadable by any skill that needs a structured slice of the Brain.

Autonomously search connected platforms (Notion/Drive/Box/Figma/Gong/Slack) for existing brand materials and produce a discovery report, what exists, where, how authoritative, and the gaps. Discovery + audit: feeds guideline generation and the voice guard.

Generate brand voice guidelines / a style guide from source documents, call transcripts, or a discovery report. Produces the artifact the voice guard enforces, claims LAW enforced throughout.

Enforce Agency Script brand voice, terminology, and claims discipline on any content. Used internally by every content skill.

The intake layer for any engine-backed job, because the brief is the interface. Validates a full brief without re-asking, asks a partial brief ONLY for its gaps, or runs the sectioned questionnaire from nothing; soft fields skip into ledgered assumptions, hard fields (identity, rights, the reference) block with a clear ask.

Query and update the 981-task Agency Script canon.

Run the self-improvement flywheel: detect a recurring manual pattern across session journals, corrections, and failures, auto-draft a candidate skill via the skill-creator, route it through the safety and curation install gates, and propose it for human approval (never auto-install).

Browse the Agency Script OS skill pack by section and pick a skill to run, the menu/directory.

Make the company get smarter over time, turn finished work, user corrections, and post-task retros into confidence-scored learnings written back to the Project Brain, and propose new/updated skills when a pattern recurs above threshold. Runs automatically when a task completes or a correction is given; promotion to a new skill is human-gated.

Strip the AI tells from any prose so it reads like a person wrote it, throat-clearing openers, emphasis crutches, jargon, hedges, and the structural clichés that survive a word swap (binary "not just X but Y" contrasts, negative listing, rule-of-three tics, fragmentation, rhetorical setups, false agency, passive voice).

Search across ALL connected sources in one query (chat (Slack), email (Gmail), docs (Notion/Drive/Box), knowledge bases (Guru), and project/issue trackers (Linear/Asana/ClickUp)) then rank and dedupe results with source attribution and a confidence level. Also manages which sources are connected.

Change existing, working work by shipping a new version ALONGSIDE it instead of overwriting it (variant registries, preview params, feature flags, snapshots) so both can be seen doing their real job before a human picks.

Capture a verified mistake as a typed anti-pattern and auto-inject its guard before a similar future task repeats the error, the loss-side counterpart to compound-learning, built to hunt the fake-success failure mode. Post-mortem the loss, type it, write a trigger-to-guard ledger entry, match its shape on the next task, surface the guard at the irreversible seam.

Write the handoff document that makes a fresh chat productive in one paste. Captures the situation in one line, what is done with its evidence, what is in flight and exactly where, what is blocked and on whom, the traps already paid for, and the single next action. Then files it to /planning so the thread survives the session.

Meta-skill that designs a domain-specific agent team and generates the agent definitions and skills they use.

Tune a running AI agent harness (Claude Code, Codex, Cursor, OpenCode) so a run costs fewer tokens, finishes faster, and is right the first time, model routing (cheap model for cheap steps, escalate on failure), token/cost optimization (slim prompt, hold pointers not bodies, deliberate compaction), context-window budgeting, parallelization (git worktrees + the cascade method).

Design AND install agent lifecycle hooks so the right action fires automatically on the right event, the "hook-able" pillar that turns a pile of skills into an operating system. Every hook is event → guard → action.

Core category hub: the pack's own engine room.

Design a durable human-approval gate, at an irreversible step (send/post/buy/transfer/sign/delete/deploy) the workflow snapshots state, notifies the owner, BLOCKS so the pending decision survives a restart, then resumes on approve/reject/edit, logging each verdict to the governance Ledger. Unattended runs STOP-and-queue; money and PII never auto-approve.

Hand it a rough prompt and what a good result looks like, and it sharpens the prompt by running it, scoring the output against your standard, rewriting, and repeating until the gains stop. Keeps the BEST version rather than the last, stops on convergence rather than a fixed count, and reports which rule fired and why.

Watches your other loops and catches what they miss, because loops drift, the rule stops being checked, inputs shift underneath, and quality degrades so gradually nobody notices until a client does.

Apply the KISS rule to anything that got too complicated, walk the Delete Ladder (delete, merge, default, defer, explain, where explain is the failure move), count the complexity budget (decisions, steps, concepts, surfaces, time-to-value), and re-architect into default/adjust/full-power layers so the surface shrinks while every capability survives underneath.

Design and build a knowledge base or documentation set for ANY situation (help center, developer docs, internal wiki, agent retrieval corpus, onboarding path, SOP library) from one source of truth that serves human readers and AI retrieval at once. Decides which builder you actually need, then owns granularity, provenance, freshness, taxonomy, and access.

Per-task model-tier routing (cheap/default/pro by role) across all client workloads, plus an LLM-COGS ledger tagged by client/project/skill with budgets, alerts, and a margin report flagging engagements where token cost eats the retainer, making AI spend a governed P&L line, not a surprise invoice. Keywords: tier routing, LLM COGS, token attribution, cost ledger, burn-rate alert, margin report.

Pick a proven, ready-to-run agent loop off the shelf instead of designing one from scratch, browse a curated library of loop recipes (docs sweep, production error sweep, flaky-test stabilizer, full product evaluation, CVE burndown, SEO/GEO visibility, pre-publish source-check, accessibility repair, and more), fill the [brackets], confirm its stop condition + verifier.

Run the periodic "sleep" pass that keeps an agent's persistent memory lean and recall sharp, read the memory store, merge duplicate notes, reconcile contradictions, prune the obsolete, compact verbose entries, and rewrite the top-level index as an index not a dump. The agent analog of sleep, sessions accumulate, the dream consolidates into durable memory.

For the rare high-stakes, one-shot decision, run a panel of diverse models instead of one, several models answer the same prompt independently, a judge model critiques and reconciles them, and the result is fused into a single answer stronger than any one model's.

Decide where large content sets (blogs, programmatic pages, docs) LIVE and keep builds fast as they scale to 10k+, database vs markdown-in-bundle vs object storage, plus the failure modes (EMFILE "too many open files", oversized ISR payloads, frozen-lockfile deploy breaks, repo/bundle bloat) and the safe files→DB migration. Sibling of as-seo-technical (which owns the rendering side).

Operate an Obsidian vault at the file/tooling layer, author valid Obsidian-Flavored-Markdown notes (wikilinks, embeds, callouts, properties/frontmatter, tags), structure the vault (MOCs, daily/atomic notes), and run bulk ops via the Obsidian CLI.

Explain and navigate how the pack works, hubs vs leaf skills, the three execution modes (on-demand / scheduled / triggered), the Project Brain context layer, the 10 guardrails, and which skill or hub to use for a given goal. Renders the operating-model diagram on request.

Turn messy saved operator language into a filled, slot-based Agency Script OS prompt that already routes to the right skill chain, continue, handoff, launch audit, homepage, logo, build loop, wow-native, and the rest of the operator shelf.

Audit, score, and improve the Agency Script OS Skills pack itself. Run the quality audit, fix trigger collisions, tighten weak descriptions, verify gate coverage, and keep the README/version in sync.

Turn a no-API web portal behind only a login (a vendor, bank, or ad console) into a normalized, scheduled data export, for AUTHORIZED own or client properties only. The governed portal-export primitive: authenticated headless fetch, named credentials never logged, MFA and transactional steps human-gated, schema-bound with change detection.

Build and maintain the perfect AI Brain for a project, a structured, agent-readable context layer (spine + curated knowledge + decisions + references) with the best guardrails on the market baked in, so every skill runs against ultimate, governed project context. Loaded by other skills that need project context.

Turn each agent session into durable, governed memory in the Project Brain, compress the session's exhaust (tool-uses, edits, decisions, open threads) into a dated, human-readable summary, redact secrets/PII at capture, file it to the Brain, and re-inject the spine + last-N sessions at the next start so the next session begins where the last one ended.

Author, upgrade, and ship an Agency Script OS skill to the house bar, description that routes, lean SKILL.md over deep canon, the ten self-sufficiency dimensions, and the mandatory baseline-vs-with-skill gate that kills skills which only tie an expert baseline.

Point at any OpenAPI spec, API URL, or Postman collection and emit a vetted connector row plus a thin skill scaffold (named-credential auth seam, rate-limit/backoff, idempotency, pagination, typed errors) conforming to pack conventions, collapsing the long-tail-connector problem into one step.

Curate an incoming community skill, MCP, or plugin into our pack, one pipeline routes safety to as-sec-skill-supply-chain, checks license + provenance, scores stack/audience FIT, dedupes capability-overlap against what we own, returns an adopt/adapt/decline verdict + placement.

Design sub-agent delegation so a long agent session stays sharp, when the main session should spawn focused, stateless workers (fresh context, own model, restricted tools) that do one scoped job and return only the distillate, keeping the main thread clean.

The pack's secret power commands, five extreme execution modes that raise HOW any skill runs, never what it may do. ultra (rubric first, parallel specialist lanes, adversarial verify, measured gates, run report), swarm (many lenses, loop-until-dry discovery), arena (competing attempts, judge panel), proof (attack an existing deliverable), ship (gates, self-score, package).

Agency Script OS startup/router and Superpowers replacement. Inventories local as-* skills, routes to the best Agency Script OS skill chain, loads Project Brain, canon, trust gates, and applies proof-first, test-first, root-cause-first, no-placeholder execution.

Chain skills into a named, repeatable, end-to-end workflow, the "workflow-able" pillar, so one input produces a finished outcome instead of re-typing the same prompts.

Design and BUILD the backend service surface, REST/GraphQL/RPC API design (contract-first OpenAPI/GraphQL, resource + verb + status-code design, consistent response envelopes, cursor pagination, versioning), end-to-end input validation (Zod), object- + tenant-level authorization (no IDOR), typed error taxonomy, idempotency + rate limiting, queued background work (BullMQ).

Standards that make an app feel fast: cached/deduped data layer instead of fetch-on-mount, navigation that never dead-ends, honest loading/empty states, optimistic mutations, and measurements. Framework-agnostic.

Design and build the async work layer beneath the API, move slow, external, or bursty work off the request into a durable job queue.

Review a diff or PR for correctness, performance, readability, and maintainability on the Next.js/TS/Prisma stack (N+1 queries, error handling, edge cases, missing tests) and route security-sensitive surface to the appsec review.

Understand an unfamiliar codebase fast, get a navigable map of how every file, function, and dependency fits together. A real parser (tree-sitter/AST) extracts the structural graph (imports, defs, calls, inheritance) as ground-truth facts; an LLM layer adds summaries, layer grouping, and code→domain mapping (labeled interpretation).

Run engineering as a compounding loop where each unit of work makes the next easier, ~80% planning + review, ~20% execution.

Design the data model + database schema BEFORE the API, model the domain (entities, attributes, relationships), draw the ERD, normalize then denormalize on purpose, pick stable keys + correct types (numeric money, timestamptz, enums), index every FK + query path, enforce integrity IN the database (unique/check/FK constraints, ON DELETE behavior).

Run a structured debugging session on the Next.js/TS/Prisma/Vercel stack, reproduce, isolate, diagnose root cause, fix minimally, and add the regression test that prevents it.

Remove code, features, endpoints, and systems safely, code is a liability and deleting it is an underrated win, but ripping it out blindly breaks consumers you forgot about.

Adversarially verify the agent's own confident output before committing, the most dangerous output is the one it's surest about.

Build an eval harness that proves a prompt, skill, agent, or workflow actually works (and stays working) turning "this seems better" into a measured score on a held-out test set.

Design and build the file-upload and media pipeline, take large uploads direct to object storage with short-lived presigned URLs instead of proxying them through the app server.

Manage Agency Script OS git workflow for isolated feature work and branch completion. Inspect dirty state without reverting user changes, choose current branch vs codex/* branch vs git worktree, stage only owned edits, make atomic conventional commits with no AI attribution, prepare PR/merge options, and clean up safely.

Engineering category hub: routes to the right leaf for building (data model, API backend, background jobs, realtime, file uploads, RAG, MCP server), changing safely (refactor, tech debt, deprecation, migration, debug, performance), and proving it works (testing strategy, webapp testing, evals, code review, verify-before-done, doubt-driven). Entry point for the engineering domain.

Design, build, and harden an MCP server so an agent can actually use it, tool surface sized to the job, schemas that fail loudly, responses shaped for a model rather than a REST client, auth and metering, and an agent-run evaluation instead of a green unit test.

Optimize application performance the measure-first way, reproduce the slow case, profile to find the real bottleneck, fix that one dominant cost, re-measure to prove the win, and guard against regression. Never optimize from intuition; the hot path is almost never where you think.

Build a retrieval-augmented generation (RAG) system that grounds an AI on your own data. So it answers from real, cited sources instead of its training priors.

Design and build the realtime layer, push live updates to the browser instead of polling, over the right transport.

Execute a safe, behavior-preserving refactor on a working app, pin behavior with characterization tests, find the seams, then run the named workflow (god-file split, duplication collapse, state untangle, boundary extraction, strangler-fig replacement, AST codemod sweep, type hardening) in small reversible steps that stay green, never mixing a refactor with a behavior change.

Ground every framework, library, or API decision in the official source instead of memory. So the agent never hallucinates a method, option, config key, or version behavior that doesn't exist.

Generate a truthful standup update (yesterday / today / blockers) synthesized from recent commits, PRs, and ticket movement, not aspirational filler.

Identify, categorize, and prioritize technical debt across the Next.js/TS/Prisma codebase into a ranked refactor backlog by impact × effort. Not a rewrite-everything wish list.

Design a test strategy/plan for a feature or the app and write the tests, the pyramid (Vitest unit/integration, Playwright e2e) tuned to this stack, coverage that matters on money/auth/render paths, and the actual test code.

A hard pre-completion gate that blocks any "fixed / passing / shipped / done / works" claim without fresh runtime evidence, matching the evidence class to the claim type, demanding it be produced this session after this change, reading the actual output instead of pattern-matching it, and labeling honestly anything that could not be verified. Attacks the fake-success pattern.

Test a running web app end to end with Playwright (drive real browser flows (signup, login, checkout, forms, dashboards), assert on visible state, capture screenshots/console/network on failure, and check responsive + basic a11y) then hand back a runnable spec + a triaged failure report. For verifying a live local/staging app behaves, not unit tests.

Inspect any digital surface to Tesla Highland fit-and-finish (gaps, paint, NVH, interior, crash structure) and return a blocker-scored card. PPL vehicles stay as-ops-lead-gen-tesla-build; chrome build stays as-ops-premium-ui-builder; design system freeze stays as-design-system-architect.

Static coherence QA for any Next.js/React app, is it NAVIGABLE, CONSISTENT, and WIRED? Runs a repeatable 8-pass scan (route inventory, internal-link integrity, nav reachability/orphan routes, design-token + CTA-primitive drift, multiline-aware form wiring, metadata/SEO coverage, static a11y, hygiene + shadow routes) and returns severity-tiered findings with file evidence.

Produce the morning growth briefing, signups, revenue, support SLA, social, hot leads, anomalies, and today's priorities.

Score, segment, and route incoming leads into the right sequence or to sales, per the lead-scoring rules.

Turn a fresh inbound lead into a qualified, enriched record plus a ready-to-send personalized reply and a booking link, within minutes of capture. Enrich → score/qualify (via as-lead-router) → draft reply → propose meeting times → log to CRM.

Turn one target account into a complete pre-call dossier, firmographic enrichment, news/triggers scan, stakeholder map, tailored angle + objection prep, and a one-page brief the operator can read in the elevator. Composes as-sales-account-research + as-sales-call-prep into a single deliverable.

Research a company or person and return actionable sales intel for Agency Script (fit, pains, openers, decision-makers).

Turn a scoped deal into the three documents that close it, a branded Statement of Work PDF, the master Terms and Conditions it attaches to, and (when the deal needs persuasion) a separate proposal or comparison talk track.

Generate tailored sales assets (pitch decks, one-pagers, proof matrices, sales scripts) from deal context.

Turn call notes or a transcript into action items, a CRM-ready summary, and a personalized follow-up email.

Prepare for a sales or partnership call, account context, attendee research, agenda, and likely objections.

One company input → branded meeting-ready discovery/audit report (HTML + PDF) with the Agency Script logo. SpaceX first-principles order (requirements less dumb → delete unused modules → simplify one shell → accelerate cycle time → automate last) plus Tesla fit-and-finish on the PDF.

The closing-call system that owns the moment of the ask, pre-close framing, the discovery to pitch to price to objection-handling to ask flow, the assumptive and summary closes, scripted objection responses for price, time, trust, and spouse/partner, and the follow-up sequence that brings a stalled deal back to a close.

Run the diagnostic discovery / qualification call (the structured first sales conversation that earns the close) with an 8-stage flow (rapport + upfront contract → situation → problem → implication → vision → MEDDIC-light qualify → reflect the gap → next step), a SPIN question bank, a ~30/70 talk-listen discipline, and honest fit/no-fit qualification. Never pressure or income claims.

Enrich a thin lead/account list the GTM-engineering way, chain data providers cheapest-first, stop on the first verified hit, merge partials, and never fabricate a field. A provider-agnostic method, not a wrapper around any one tool.

Build a tailored enterprise proposal/SOW for Agency Script (team certification, cohorts, enterprise assessment, capability reporting) inside the hard enterprise tier boundaries (no white-label, no IP transfer, read-only dashboards).

Build a weighted pipeline forecast from a CRM export or pipeline CSV, best/likely/worst cases, commit vs upside, and gap-to-quota with the deals that close it.

Sales category hub: routes to the right leaf for building the list (ICP from signal, TAM, prospecting, enrichment), opening (outreach, multichannel cadence, call prep, account brief and research), running the call (discovery, methodology, closing, follow-up), the assets (pitch deck, proposal, presentation, discovery report, sales asset), and managing the number (pipeline review, forecast).

Watches the pipeline every day, finds the deals that have gone quiet, drafts the specific follow-up each one needs, and queues it for your approval before anything sends. Catches the deals that die from being forgotten rather than from being lost, which is most of them.

Map the operations-engineering-finance buying committee and ROI math rules for industrial deals. General qualify lenses stay as-sales-methodology.

Build a qualified prospect LIST and turn it into outreach (discover local/niche businesses by category + location + radius (grounded in Google Maps/Places data) website, phone, rating, reviews), qualify and tier them on real need signals (no/weak website, low reviews, marketing gaps, ICP fit), run a quick audit on the hot ones to find a specific pitch hook, then draft personalized.

Qualify and strategize a deal with the right named sales methodology, MEDDPICC, SPIN, Challenger, Gap Selling, Value Selling, and Sandler. Picks the lens for the job, runs it over the live deal, and surfaces the gaps as a prioritized next-step action list, e.g. a MEDDPICC red/yellow/green scorecard whose reds are the work.

Author ONE cross-channel outbound cadence (email + LinkedIn + SMS + voice as a single sequenced object with reply-branching, per-channel consent, and one shared exit) that compiles down to the per-channel executor skills. The conductor the single-channel skills lack, as-email-cold-campaign owns the email sequence, this owns where every channel sits in one arc.

Draft personalized cold and warm outreach (email, LinkedIn, follow-ups) for Agency Script prospects and partners.

Run a pipeline-health review: prioritize the deals that matter, flag stale / single-threaded / bad-close-date deals, and produce a weekly action plan.

Turn one deal/offer context (or raw proposal text) into a branded, story-driven pitch/proposal deck, audience+pain framing, story arc, canon-backed proof matrix, pricing/offer slide, rendered to pptx/Gamma/Canva (or a deck spec if no connector).

Build or AUDIT any persuasive presentation on the 4-Frame buyer-sequencing spine (Why → What → How → Now) so it moves the audience to a decision instead of dumping information, sales decks, webinars, keynotes, VSLs, investor pitches, internal-strategy decks, demos, discovery calls.

Deliver the paid AI Readiness Consultation end to end (intake analysis and pre-scoring, the 4-dimension readiness rubric (operations, delivery, data, governance, 0-100 with evidence lines), the 60-minute live-call runbook, the 48-hour findings report, and the credit-to-program ascension mapping) paid discovery run with consulting-firm rigor, never a disguised pitch.

Discover your ICP from outcome data instead of opinion, mine the real difference between closed-won and closed-lost (or churned) accounts to find the signals that separate winners from losers, then build a transparent 0-100 fit-scoring model that ranks new prospects.

Turn an ICP into a real, worked account+contact list, the GTM-engineering "build a TAM" play, provider-agnostic and compliant.

Mines the exact words your customers use from sales calls, support tickets, reviews and won-deal threads, then feeds that language back into ads, landing pages, emails and scripts so your marketing stops sounding like you wrote it and starts sounding like your buyers did.

Surface the assumptions an idea rests on, rank them by impact times uncertainty, and test the riskiest leap-of-faith ones with the cheapest experiment before committing build effort.

A sharp product thinking partner, explore a problem space, challenge assumptions, and sharpen a fuzzy idea into a crisp problem statement plus the next cheapest experiment. This is everyday product judgment (divergent→sharper problem); for convergent strategy/moat/billion-dollar bets use as-ops-strategic-advisor, and for full feature specs use as-pm-write-spec.

Product management category hub: routes to the right leaf for deciding what to build (brainstorm, assumption mapping, market validation, opportunity solution tree, strategy frameworks, requirements physics), setting direction (North Star, OKRs, roadmap), understanding users (journey map, synthesize research), and shipping (write spec, sprint planning, pre-mortem, stakeholder update).

Map the end-to-end experience a specific persona has getting a job done (the ordered stages, and at each what they do, the touchpoints, what they think, how they feel, and where it hurts) built from real research, not invented at a desk. Plots the emotion curve, marks the lows and the moments of truth (first value almost always one), and turns each significant pain into a prioritized opportunity.

Validate whether the market will actually pay for a new offer/product/service line BEFORE you invest in launch.

Define the North Star Metric, the single number that captures the core value your product delivers and leads revenue, not a lagging output or a vanity count.

Set OKRs that create focus, not a task list (Wodtke, Radical Focus), one ambitious qualitative Objective per team per quarter plus 2-4 measurable outcome Key Results with baseline to target and a data source, laddered to the company objective and the North Star.

Build an Opportunity Solution Tree (Teresa Torres, Continuous Discovery Habits), the visual map from one measurable outcome down through research-grounded opportunities (customer needs/pains/desires in their words), to 3+ compared solutions per target opportunity, to the experiments that de-risk them.

Run a pre-mortem (Gary Klein) and red-team a plan, assume the project already failed and work backward to explain why, which defeats optimism bias and surfaces the risks people stay quiet about.

Strip a brief to physics-only requirements before the PRD is written. Full PRD is as-pm-write-spec; full Algorithm including delete is as-ops-the-algorithm.

Build, update, and reprioritize the product roadmap in Now/Next/Later horizons with RICE or value-effort scoring and explicit swap tradeoffs. For big end-to-end feature builds use as-ops-product-feature-builder; for strategy/moat/billion-dollar bets use as-ops-strategic-advisor; for KPI reporting use as-analytics-*.

Plan a sprint from real capacity to a single goal, PTO/meeting/carryover/buffer math, P0 vs stretch, and owner-runs-irreversible flagging. For the durable roadmap use as-pm-roadmap; for big feature builds use as-ops-product-feature-builder; for KPI reporting use as-analytics-*.

Write an audience-tailored stakeholder update (exec brief, eng detail, or customer-facing) in status, launch, or escalation flavor. For brutal strategy use as-ops-strategic-advisor; for KPI/metrics reports use as-analytics-weekly-metrics; for full internal-comms formats use as-design-internal-comms.

Run the right strategy framework for the question and fill it with honest analysis, not box-filling, Business Model Canvas, Lean Canvas, Value Proposition Canvas (JTBD), SWOT, PESTLE, Porter's Five Forces, Ansoff Matrix, and Product Vision. Picks the framework by the question, then runs it so every cell ends in an implication or action and the analysis points to one decision.

Synthesize existing product feedback (interviews, surveys, support tickets, NPS, cancellation reasons) into themes ranked by frequency × impact, then into roadmap recommendations. For PLANNING and RUNNING UX research (interview guides, surveys, usability tests) use as-design-user-research. This one turns existing feedback into PM decisions.

Turn one problem statement or feature idea into a complete, build-ready PRD file landed in /planning (framed problem + segment, goals/non-goals, operator user stories with testable acceptance criteria, 1-3 sourced success metrics, and a phased rollout) ready to hand to as-ops-product-feature-builder.

Own CAUSAL, incremental marketing measurement, choose the model by a spend/data-maturity tier (small → lift tests; mid → multi-touch attribution checked against holdouts; large → Marketing Mix Modeling + geo-holdout incrementality), never defaulting to the fanciest, and separate a real mix shift from a tracking artifact via signal-integrity diagnostics (Consent Mode v2, server-side tagging, CAPI).

Produce the white-labeled monthly client report from one input (client name + connected accounts), pull GA/GSC/Meta/ads metrics, trend WoW/MoM, narrate wins/misses, set next-month priorities, render under the client's logo and brand. The #1 recurring retainer time sink, done in one pass.

Build a live, reopenable executive dashboard (HTML artifact) that pulls fresh numbers from connected tools on each open.

Configure and AUDIT Google Analytics 4, read via the official read-only analytics MCP, write config via the GA4 Admin API (mark/create key events + conversions, custom dimensions/metrics, enhanced measurement, data streams, retention), audit configured-vs-actually-firing events, and apply the 10/10 conversion blueprint for lead-gen/agency accounts.

Analyze a campaign, channel, or content set and report wins, misses, and prioritized optimizations.

Compose owned analytics (Umami for privacy web stats, PostHog for product/flags/replay, GlitchTip if Sentry residency hurts) one tag per job, licensed, on hostssh. Not GA4 admin (as-analytics-ga4-admin), not a monthly client report (as-analytics-client-monthly-report).

Produce the weekly business metrics report (acquisition, activation, revenue, retention, content/SEO, and pipeline) with trends and recommended actions.

Answer a data question end to end, from a quick lookup to a full investigation (explore → query → analyze → validate → visualize → communicate).

Capture a company's data-warehouse tribal knowledge into a reusable reference, schemas, metric definitions, join patterns, gotchas, and the canonical query for each KPI.

Turn a scanned or photographed document into governed, structured JSON, layout-and-table-aware OCR for signed contracts, receipts, faxed forms, and flattened legacy PDFs, with per-field confidence, source bounding boxes, internal-consistency validation, and a low-confidence human-review queue. Never fabricates a field; PII is handled, never leaked.

Profile and explore a dataset before analysis, shape, grain, nulls, distributions, cardinality, duplicates, outliers, and data-quality issues.

Data category hub: routes to the right leaf for answering a question (analyze, explore, SQL, stats), trusting the answer (validate, lineage and provenance), showing it (viz, JSON canvas), and standing up the sources (owned corpus, knowledge graph, document OCR extraction, warehouse context). Entry point for the data domain.

Turn an idea, structure, or set of notes into a visual JSON Canvas (.canvas) file, mind maps, flowcharts, project/kanban boards, and research canvases, valid to the JSON Canvas 1.0 open spec (nodes + edges), grid-aligned and grounded in real source.

Turn a pile of source documents (PDFs, markdown, transcripts, scrapes) into a governed, human-readable entity-relation knowledge graph in the Project Brain, extract → profile → dedup → index → dual-level query.

Trace every client-report number back through its pipeline (source feed → join → derived column → metric → report cell) and fingerprint each upstream feed so a silent shape change is caught before the report ships, not after the client asks. Keywords: lineage, provenance, source-of-record, shape drift, schema drift, fingerprint, blast radius, impact analysis, fanout, attribution-window.

Stand up and keep fresh the first-party corpus every content system runs on, credential-safe routines that pull call and meeting transcripts, YouTube and podcast transcripts, every email you have sent with its open and click stats, wins screenshots, and case-study interviews into one governed folder with stable filenames, incremental sync, and duplicate-recording dedupe.

Write correct, performant SQL for any warehouse dialect. Or fix, optimize, or translate an existing query.

Run statistical analysis (descriptive stats, trend/seasonality, outlier/anomaly detection, hypothesis tests, and correlation) with the right method and the misuse caught.

QA an analysis before it's shared, check methodology, accuracy, denominators, timezones, filter leakage, and bias.

Build a publication-quality chart: the right chart type for the question, accessible and honest-axed (matplotlib/plotly).

Per-engagement research operations desk: decompose the question, gather independent sources, cross-source triangulate, cite or flag every claim, then synthesize a verified dossier (company · market · competitors · regulatory · buying-signals) and refresh it on a schedule. Productizes the $5-15k research deliverable.

Photoshoot-free product creative for DTC/apparel, one flat shot becomes a cutout, on-model/try-on renders, lifestyle scenes, and per-marketplace exports (Amazon main on white, Shopify, Etsy), each passing a truthfulness QA so the product shown is the product sold. Single-hero/brand assets go to as-design-*; ad concepts to as-ads-meta-creative.

ICE repair, collision, tint, tires, mobile mechanics, shop economics and trust deficit. Canon: VERT_AUTO_SERVICES.md.

Salons, barbers, and non-clinical spas, review+photo flywheel, rebooking, chair utilization. Clinical med-spa claims leave this skill.

Vertical overlay for commercial/industrial solar, storage, demand-charge, microgrid GTM. Residential stays as-vert-energy-home; RFPs stay as-ops-grants-rfp.

Vertical overlay for residential solar, battery, and EV-charger installers, economics, claims, backlog, 90-day plan, then route to domain skills. HVAC/handyman stays as-vert-home-services; C&I stays as-vert-energy-cni.

Vertical overlay for EV retail/conversion, fleet, and charge-point hosts. Not ICE repair shops. Oil-change/collision stays as-vert-auto-services.

Gyms and studios: membership density, class pack vs contract, no body-shame, no fake before/after. Canon: VERT_FITNESS.md.

Public commercial GTM for climate-tech, robotics, and dual-use-adjacent companies with a hard refuse on ITAR, classified, and weapons. Grants stay as-ops-grants-rfp.

The home-services vertical umbrella (load the shared playbook for ANY local residential-service client (handyman, multi-trade contractor, any trade without a dedicated pack), diagnose the business, then route execution to the domain skills with vertical parameters) homeowner trust psychology, diagnostic-fee/membership offers, map-pack dominance, review velocity, LSA/PPC, speed-to-lead.

Vertical category hub: routes to the industry overlay that adapts the growth machine to one trade's economics, seasonality, buying committee and advertising rules.

Landscaping and lawn: seasonal capacity, recurring routes. Canon: VERT_LANDSCAPING.md.

Law firms: ethics advertising rules; this overlay never renders legal conclusions. Canon: VERT_LEGAL_FIRMS.md.

Vertical overlay for job shops and contract manufacturers, market empty capacity or higher-value work. Inventory math stays as-ops-supply-chain.

HIPAA/ADA-aware clinics and dental. No diagnose/cure copy, no review gating, treatment-plan economics. Canon: VERT_MEDICAL_DENTAL.md.

Pest control: recurrence contracts, chemical/label claims gated, seasonal swarm windows, not a generic trade. Canon: VERT_PEST_CONTROL.md.

Agents, teams, brokerages: Fair Housing binds every asset. Canon: VERT_REAL_ESTATE.md.

Restaurants and food service: offer-led, geo-tight, no fake covers/hour. Canon: VERT_RESTAURANTS.md.

Roofing and exterior contractors: storm ethics, insurance-claims language gated. Canon: VERT_ROOFING_EXTERIORS.md.

Licensed electricians, plumbers, HVAC: ticket mix, after-hours, permit jobs; extends home-services but is not the umbrella. Canon: VERT_SKILLED_TRADES.md.

The margin bridge: take the per-client LLM-COGS ledger, turn metered AI usage into a billable / margin line item on client reporting under the agreed billing model (pass-through / markup / bundled), and flag engagements where token COGS erodes the retainer; a human approves every invoice.

Chase overdue invoices and recover the cash you are owed (an accounts-receivable recovery loop for HUMAN-SENT service invoices) pull AR aging, segment by amount/risk, run the escalating dunning ladder (friendly to firm to final + payment-plan), stop on pay, trigger a service-hold when they still will not.

Cap-table hygiene and scenario modeling prepared for counsel, every holder, instrument, share class, vesting schedule, and exercise status reconciled against the signed documents (the source of truth, drift flagged), SAFE/note conversion math (caps, discounts, pre/post-money, MFN) as stated-assumption scenarios, dilution with the option-pool shuffle visible, pool refresh as board questions.

Manage the month-end close. Build and run the close calendar (business-day-offset task list), sequence tasks by dependency, track status and the critical path, and surface blockers. Formal accounting close: ties to OPERATIONS_RUNBOOK.

Build the forward-looking finance layer (driver-based operating models (revenue drivers, headcount, opex), 3-statement linkage, rolling 12-18-month forecasts, the budget cycle, base/upside/downside scenarios with named swing drivers, cash runway math, and the monthly reforecast against actuals) with agency (utilization, bill rates, project margin) and SaaS (MRR waterfall, NRR, CAC payback).

Finance category hub: routes to the right leaf for keeping the books (pre-accounting, journal entries, reconciliation, close management, statements), looking forward (FP&A modelling, variance analysis, cap table, tax strategy), getting paid (AR collections, AI COGS to invoice), and surviving scrutiny (SOX control testing, audit support). Entry point for the finance domain.

Prepare GAAP journal entries (accruals (AP, payroll, prepaid), depreciation/amortization, and revenue recognition / deferred revenue) balanced (debits = credits) with full supporting detail for audit. Drafts for human posting; never posts autonomously.

Cut cost by deleting parts and changing design, using their invoices as the dollar range. Idiot index is the ratio; FinOps is the cloud bill.

The shoebox-to-clean-books loop: capture receipts and invoices, OCR them, match against Stripe and the bank feed, categorize, then draft balanced journal entries a human reviews and posts; what won't match queues as an exception. The pre-accounting orchestration the finance pack assumes done. Never posts or moves money.

Reconcile accounts (bank (GL cash vs statement), GL-to-subledger (AR/AP/revenue control accounts), and intercompany) and classify every reconciling item as timing / error / unidentified. Reconciliation flags discrepancies; it never moves money.

Run SOX 404 control testing and audit support, define the population, select a defensible sample, perform the test, build the workpaper, and classify deficiencies (deficiency / significant deficiency / material weakness). Drafts conclusions for a qualified human to sign.

Generate GAAP financial statements (income statement / P&L, balance sheet, and cash flow (direct or indirect)) with period-over-period and budget comparison, variance columns, and tie-outs. Formal corporate accounting: distinct from as-ops-finance-ops (SaaS MRR/billing).

Tax-season readiness prepared for the licensed professional who decides, entity trade-off questions for the CPA (sole prop / LLC / S-corp), deduction/credit documentation inventories (home office, vehicle, R&D readiness, retirement), timing worksheets, estimated-payment calendar + safe-harbor concepts, multistate/nexus flags, and the year-end preparer package.

Decompose budget-vs-actual variances into drivers (volume × rate, mix, timing, one-time), build a reconciling waterfall budget→actual, and write an on-brand narrative. Formal FP&A: distinct from as-ops-finance-ops revenue snapshots.

Build a person's own AI "clone", a coordinated Brain + Image + Voice + Video set that produces on-brand content at scale so the owner serves instead of grinding out posts. Assembles the Brain (Positioning Context Library), Image clone (AI photos), Voice clone (TTS), and Video clone (talking avatar) into one automated content super-agent, then routes it into proven viral formats and the funnel.

Build a long-form article as a designed reading experience, measure and rhythm, a typographic scale that earns its display sizes, pull quotes and figures that carry weight, code and data set properly, progressive disclosure of depth, reading progress, and a version that survives being read on a phone in bad light.

Produce a complete, SEO-optimized, on-brand Agency Script blog post end to end. Avoids duplicating the existing 1,783-post library.

Plan and maintain the Agency Script content calendar across blog, social, email, and video.

Write build-in-public engineering narrative: hardware, test, failure, next flight. Founder guru essays stay as-content-thought-leadership-essay.

Content category hub: routes to the right leaf for planning (editorial calendar, intelligence scanner), writing the flagship piece (blog engine, thought-leadership essay, engineering narrative, press release, media bio, video script), multiplying it (repurposer, podcast repurpose), and producing shows (podcast producer, AI clone builder). Entry point for the content domain.

Your strategic content radar (scan the external sources you already follow (newsletters, RSS, Substack, subreddits, X bookmarks, saved links, competitor blogs) whatever you name) and return publishable content ANGLES grounded in your contrarian stances and voice, not summaries and not generic takes. Every angle traces to a real cited signal AND one of your positioning stances.

Write a compelling professional bio that positions you as the go-to expert, in three ready-to-use lengths (one-liner, short ~50-word, and long ~150-200-word) plus first- and third-person versions, for speaker intros, podcast guesting, author/about pages, social profiles, and press/media kits.

Write or rewrite memorable stories with the ONION framework: Objective, Novelty, Imagery, Obstacle, Nugget. For ad, video, essay, or email work, this skill builds the narrative spine before the channel-specific skill formats it.

Plan, produce, and publish podcast episodes end to end, show positioning, the episode pipeline (plan, record-prep, edit brief, show notes, publish), trend and topic scouting, guest research with human-gated outreach, plus podcast SEO across titles, chapters, and transcript.

Turn one recording (podcast, Zoom, or webinar) into a full repurposing pack, transcript, chapters + highlights, show notes with title options, a timestamped clip pick-list, and a channel-native social set + newsletter.

Write a press release journalists will actually run, built on the proven 4-step method (identify the ≤27-word news hook → answer the 5 W's + H and gather 2-3 real attributed quotes → order facts inverted-pyramid with the brand message woven in → QA against a coverage checklist).

Turn one source asset (blog post, essay, webinar, call) into a full multi-channel content set, LinkedIn posts, X threads, newsletter, short-form video scripts, and image briefs.

Write a high-authority point-of-view essay that advances the Agency Script thesis on AI capability assurance and standards.

Write video scripts (YouTube long-form, short-form (Reels/Shorts/TikTok), course lessons, and demo walkthroughs) in the Agency Script voice.

Build a presentation as a web page rather than a slide file, a paced, navigable, linkable deck that plays in a browser, with speaker flow, keyboard and remote control, presenter notes, a printable path and a version that reads on its own after the meeting. Owns the web-native deck; for a PPTX or PDF deliverable use as-sales-presentation-builder.

Turn one brand/topic + a target month into a complete, scheduler-ready social content pack, pillar themes, 20-30 channel-native posts (via as-social-post-factory), hooks + CTAs, image briefs, and a day-by-day schedule pushed to a poster connector (connector-optional).

Triage and draft replies to social mentions, comments, and DMs across platforms as part of the daily engagement block.

Generate a batch of channel-native social posts (LinkedIn, X/Twitter, Instagram, Facebook) from the Agency Script template library.

Compose a social graphic as editable layers (stock plate (Pexels via the house render service), AI still, type, position, font) then optionally place it. Not copy cadence (that is as-social-post-factory) and not Meta ad batches (as-ads-meta-creative).

Produce a contextual legal briefing (a daily legal scan, focused research on a legal question, or an incident briefing) that classifies matters and routes actions. Briefing + routing. Not legal advice.

Turn won-deal terms (scope, price, term) into a drafted, sign-ready client agreement, picks the right template (MSA / SOW / retainer), fills the terms, runs a risk pass, applies a brand/clarity pass, and stages it for e-signature. Drafts for counsel + the owner signs. Never auto-sends or auto-signs.

Run the contract portfolio as an operation, intake triage, drafting from the clause library (position/fallback/walk-away), negotiation redline tracking, human-gated e-sign, obligation extraction (SLAs, payment terms, auto-renew windows), a renewal calendar that beats notice deadlines, repository hygiene, and portfolio metrics.

Run the privacy program as an ongoing operation (GDPR/CCPA(+CPRA) as working process) DSR intake→identity-verify→fulfil inside the legal clock, ROPA upkeep, consent + cookie audit, DPA/sub-processor tracking, privacy-policy currency, 72-hour breach-clock readiness.

Legal category hub: routes to the right leaf for incoming paper (NDA triage, contract lifecycle, vendor check, client agreement), assessment (legal risk, data privacy programme, IP and trademark readiness), and communication (legal brief, meeting briefing, templated response). Everything is prepared for a qualified lawyer, never filed or signed.

Get IP and trademark READY for the attorney, knockout clearance prep (identical + confusable-mark sweeps across registers plus common-law web/social/domain checks), filing-readiness package (mark variants, candidate classes as questions, specimens, first-use evidence), the IP inventory with contractor-assignment gaps flagged, the renewal calendar, and infringement-signal intake.

Prep a briefing for a legal-relevant meeting (a contract negotiation, a board meeting's legal items, or a compliance review) and track the action items out of it. Prep + action-tracking. Not legal advice.

Triage an incoming NDA fast, classify GREEN (sign-ready) / YELLOW (negotiate) / RED (escalate), with an embedded non-solicit/non-compete/missing-carveout scan. Prepares for counsel. Never auto-signs.

Draft a response to a common legal inquiry (data subject request (DSR), litigation hold, vendor legal/security questionnaire, or subpoena) with built-in escalation guards. Drafts for counsel. Never auto-sends.

Assess and classify a legal risk (a contract term, a deal, a feature, or an incident) on severity × likelihood with an explicit escalate-or-not call. Prepares for counsel. Not legal advice.

Report the status of all agreements with a vendor across systems (Lumin/Box/Drive/Gmail) and run a gap analysis (MSA/DPA/SOW). Status + gaps: remediation/signing is owner-run.

Build a web page or prototype at award-tier craft in one directed pass, verify facts, set five art-direction dials, offer three differentiated directions instead of an interview, declare the design system before any code, show a v0 draft early, then full build and verified handoff.

Build a playable browser game end to end, loop and camera, level authoring and map editing, enemy and monster systems, combat feel, encounters and tuning, inventory, audio feedback, VFX, rigged and hybrid assets, mobile performance, playtesting and shipping. Consolidates the full game lane into one directed pass.

Scaffold and maintain the Divi child theme foundation, design tokens, functions.php enqueue patterns, custom ET_Builder_Module registration, and the asset pipeline.

Build the rare, self-contained interactive widget for a Divi Code Module (calculators, custom animations, embeds), last resort only, with a registry so widgets don't multiply silently.

Generate importable Divi Library/Portability JSON layouts so pages stay natively editable in the Visual Builder, with validation before import.

Web build category hub: routes to the right leaf for WordPress and Divi client delivery (project setup, child theme, code module, importable JSON layouts, performance and accessibility gates), commerce (Shopify storefront delivery), and modern web surfaces (page from a reference URL, scroll-driven narrative sites, PWA, internationalisation). Entry point for the web-build domain.

Internationalize and localize a web app or site end to end, externalize strings into an ICU MessageFormat catalog (plurals, gender, select), run the extract→translate→review→merge workflow behind a review gate, wire locale routing with hreflang and multilingual SEO, handle RTL/bidi, and format dates/numbers/currency via Intl.

Build or rebuild a landing page end to end, intake, one-action structure, archetype choice matched to awareness stage, conversion copy, proof placement, FAQ, SEO/AEO decision, then the craft floor, content realism and the ship checklist. One offer, one audience, one primary action.

Turn a reference URL (or a supplied design) into an import-ready WordPress/Divi page in one pass, scrape the structure and brand, rebuild it as validated Divi Library JSON, run the performance/a11y/SEO Definition-of-Done gates, glue any reusable styling into the child theme, and package it with import + verification steps.

Turn a web app into an installable, offline-capable Progressive Web App, web app manifest, service-worker lifecycle, caching strategies (cache-first, network-first, stale-while-revalidate), an offline fallback, add-to-home-screen installability, push notifications behind a consent gate, and the update/versioning flow (skipWaiting pitfalls), Next.js-aware with a Lighthouse PWA checklist.

Build a scroll-driven narrative site or video-to-website page, scroll is the timeline, acts change state under the reader's hand, with moving assets generated through whichever video provider is actually connected (CoreReflex, Higgsfield, Veo3, or supplied footage). Enforces a grammar choice, a fingerprint gate against your prior builds, and verification on the rendered page.

Build a client's Shopify storefront the Online Store 2.0 way, JSON templates, sections and blocks, Liquid, metafields and metaobjects, custom sections, plus the call on when to ship an embedded app (App Bridge, theme app extensions) instead of a theme edit, closing on a performance, a11y, and handoff delivery checklist.

Set up or pick up a custom WordPress + Divi client build, generates the project CLAUDE.md, enforces the three-path build layering (child theme foundation, Divi JSON page assembly, Code Module widgets), and defines the workflows.

Run the Definition-of-Done gates on a WordPress/Divi page, performance budget (LCP/CLS), WCAG 2.1 AA accessibility, responsive checks, and SEO basics.

Design an AI appointment-setting agent (chat or voice) that books qualified prospects straight onto a calendar (persona, booking conversation flow, light qualification, real-time availability lookup, confirmation and reminder logic, and CRM capture) then emit a deploy spec. The skill designs and configures; the live agent runs on the owner's MCP plus a chat or voice provider.

Design a website chat agent that qualifies inbound visitors and books the qualified ones (persona, qualifying conversation flow, lead scoring, knowledge grounding, calendar booking, and CRM capture) then emit a deploy spec for a web chat widget. The skill designs and configures; the live widget runs on the owner's MCP plus a chat embed on the client site.

Drive a real computer or browser safely to automate a desktop/GUI task, our own governed operating layer over any computer-use runtime (native tools, a hosted API, or any desktop-control MCP), never a dependency on one.

Design an AI executive-assistant agent that manages the owner's calendar, inbox, and to-dos from a chat channel (Telegram/Slack/email) (reads the calendar and email, drafts replies in the owner's voice, schedules/reschedules events, builds and updates a to-do list, and surfaces a daily brief) then emits a deploy spec.

Deployed-agent category hub: routes to the right leaf for customer-facing agents (chat lead qualification, support, voice, appointment setter), personal agents (executive assistant, second brain), and the rails they run on (computer use, scheduling orchestration, payment rails). These design a deployable agent spec, not a live agent.

Safe agent transaction rails: issue a single-use, merchant- and amount-locked, time-boxed virtual card inside a HARD, human-set, revocable spend envelope; in-envelope micro-charges auto-authorize, above-threshold or off-allowlist spend routes to a human, every authorization is ledgered. Transacts WITHIN a pre-authorized envelope. Not autonomous spend, never money movement (G8).

Cross-party booking broker: merges free/busy across the agency, client, and CRM calendars (Google, Outlook, CalDAV), then books atomically via a two-phase commit (hold → confirm → release) so a slot is never double-booked. The atomic-booking infrastructure beneath the as-agent-appointment-setter conversation.

Scaffold a personal executive-assistant / second brain for Claude Code, a governed context layer (CLAUDE.md + @-imported context files + rule files + an append-only decision log + a backlog) that makes the assistant know you and get smarter without re-explaining.

Design an AI customer-support agent grounded in the client's own docs and knowledge base, persona, support conversation flow, retrieval grounding, ticket-deflection logic, and clean escalation to a human with full context. The skill designs and configures; the live agent runs on the owner's MCP plus a chat (or voice) channel.

Design, script, and configure an inbound or outbound AI voice agent and emit a provider-ready deploy spec, persona, conversation state machine, knowledge grounding, voice selection, guardrails, and escalation-to-human. The skill produces the design + config + deploy spec; the live agent runs on the owner's MCP plus a voice provider (Twilio/Vapi/Retell plus ElevenLabs).

Design event-driven automation workflows (Make.com / Zapier / native) for lead capture, content publishing, social posting, lifecycle triggers, and ops.

Write high-converting SINGLE broadcast emails on a named, proven framework, Cash Injection (drive sales now), Story Seller / Soap Opera (story-to-sale), Seinfeld / Movie Sequence (daily "about nothing"), Situation / Observation (turn any moment into an email), Problem-Agitate-Solve, and Super Ticket (premium sell).

From one ICP + offer, produce a complete cold-outreach campaign, list/segment strategy, a 4-6 step sequence with full copy, branching/exit logic, subject-line A/Bs, a deliverability checklist, and the automation spec to wire it. The net-new cold engine (existing email skills are warm-lifecycle + newsletter).

Audit and fix email deliverability so campaigns land in the inbox, not spam, the foundation every email skill depends on. Covers domain authentication (SPF, DKIM, DMARC→reject, BIMI, a sending subdomain), sender warmup + reputation, list hygiene (hard-bounce removal, a sunset policy, double opt-in), engagement segmentation, a spam-trigger audit, and monitoring (Postmaster, DMARC, seed tests).

Email category hub: routes to the right leaf for writing the send (broadcast, newsletter engine, cold campaign, SMS), building the machine (lifecycle sequences, automation spec, idea mining), and making sure it arrives (deliverability audit, owned self-hosted sending). Entry point for the email domain.

Mine the business data you already own (coaching and client calls, video transcripts, your sent emails with their open and click stats, wins, and case-study interviews) into a ranked, sourced email idea ledger.

Build or refresh full lifecycle email sequences, welcome, onboarding, sales, abandoned cart, win-back, failed payment, referral, partner, podcast pitch.

Write the weekly Agency Script newsletter from the week's content, wins, and one teaching idea.

Pick and spec owned email, listmonk for broadcasts, Postal for transactional MTA, Mailu/docker-mailserver only if they accept mail-admin load. Not sequence copy (as-email-lifecycle), not deliverability audit of a hosted ESP (as-email-deliverability).

Write a compliant, high-response SMS/text-message campaign or sequence and spec its cross-channel automation, for high-intent, opted-in moments only (speed-to-lead, abandoned cart, appointment reminder, challenge/live-event, launch/cart-close, win-back), layered WITH email as one journey.

Design and operate the affiliate program, commission economics engineered against real LTV/CAC (recurring vs one-time, cookie windows), honest tracking + attribution, affiliate recruitment and vetting, claims-clean swipe kits, FTC disclosure compliance as a binding gate on every asset and agreement, fraud detection (self-referral, cookie stuffing, trademark bidding, coupon leakage).

Design and operate the managed channel partner program (resellers, agencies, SIs, and referral partners run as a revenue channel) partner ICP + tier model with promotion/demotion rules, recruitment pipeline and partner-P&L pitch, enablement (onboarding, certification tie-in, deal kits, MDF), deal registration + channel-vs-direct conflict rules, partner economics, and the QBR health cadence.

Community and partnerships category hub, routes to the right leaf for running members (community manager), programs that pay others to grow you (affiliate program, referral engine, channel partners, influencer ops, partnership outreach), and earned attention (PR and media, reputation management). Entry point for the community domain.

Run influencer partnerships as a managed program, vetting on engagement quality over follower count (fake-following triage), negotiation prep with the organic-vs-paid usage-rights licensing trap priced, briefs that free the creator voice inside guardrails, FTC disclosure as a binding gate in the agreement AND at publish, whitelisting handoff to ads, G8 milestone payments.

Run the member community: prompts, weekly programming, engagement responses, and health tracking.

Recruit and manage partners: co-marketing, channel, white-label, and integration partners.

Run PR and earned media, media kit, press releases, podcast pitches, HARO-style responses, and award submissions.

Run the referral and affiliate program, identify advocates, trigger asks at the right moment, draft referral/affiliate copy, and track attribution.

Manage online reputation and respond to a brand/PR crisis, honestly. Monitor reviews + brand + AI-answer-engine mentions, respond to negative reviews promptly and calmly (own what's true, fix the root cause), and earn REAL reviews systematically.

Run a won client account from kickoff through delivery, reporting, renewal, and growth, the agency retainer/project engine.

Turn a member or client win into a structured, on-brand case study with quantified results.

Turn retention from monitoring into modeling, churn definitions done right (logo vs revenue, the denominator law), cohort retention triangles, leading indicators discovered from usage/engagement/support/billing signals, health scores validated against actual churn, survival-curve reading, risk-triggered intervention plays with save-offer margin math, and holdout-proven win-backs.

Turn a won client deal (name + scope) into a complete, client-ready onboarding packet, intake form, kickoff agenda + timeline, access/asset checklist, comms cadence, and a welcome email + portal doc, all brand-voice gated.

Customer success category hub: routes to the right leaf for landing the client (client onboarding kit, member intake, onboarding engine), running the account (account management), keeping them (retention monitor, churn science), and turning results into proof (case study builder, testimonial collector). Entry point for the customer-success domain.

Turn a completed new-member intake form into a Member Profile (the segment (runway x capacity x commitment x support x skill x readiness), an honest churn-risk band, a curriculum entry point, and a 90-day success plan anchored to the member's own goals (never an income promise)) plus a voice-of-customer + attribution extract. The read that personalizes onboarding before it starts.

Drive new-member onboarding to first value and first credential, welcome flow, activation nudges, and the proof loop.

Detect churn risk and trigger save plays, health scoring, at-risk lists, win-back.

Systematically collect, format, and route testimonials and reviews from happy members.

Launch a branded faceless YouTube channel, tier path, channel identity, settings hygiene, production pipeline, and the first-10-videos plan.

Run a faceless YouTube channel week to week, publishing QC, analytics readout (avg % viewed, outliers, CTR), monetization mix, monthly report, and flop autopsies.

Orchestrate the faceless YouTube business system end to end, niche bending, script bending, AI + team automation, channel launch, ops, and exit.

Faceless YouTube category hub: routes to the right leaf for choosing the niche (niche hunter), launching (channel launch), producing (script bender, thumbnail, production team), running it week to week (channel ops), the whole business system (faceless YouTube orchestrator), and the exit (portfolio exit). Entry point for the YouTube domain.

Find and validate faceless YouTube niches via niche bending (proven format × untouched market) scored against the $10k/mo checklist, with a niche database and bend ideas.

Value, package, and sell a faceless YouTube channel, exit multiples, sale prep (P&L, SOPs, transferability), listing strategy, and the portfolio flip-and-relaunch loop.

Build the high-ROI production team for a faceless channel, role-by-role hiring funnels (job post, form, paid test task, QC rubric), global-market rates, training, and pipeline management.

Write viral faceless YouTube scripts via script bending, extract the skeleton from a proven script, fill it with the new niche's idea, gate against AI-slop/inauthentic/repetitive checks.

Design high-CTR YouTube thumbnails (and Shorts covers) (one big visual idea, an emotional face or hero subject, 3-5 words of huge legible text, high contrast, and a curiosity gap that pairs with the title) produced as concepts + the final artwork at 1280×720 (16:9) safe-zoned, plus 2-3 A/B variants to test. Built on real thumbnail-CTR principles, original artwork only.

Design and operate a real support/helpdesk function, the tier model, SLA/SLO policy, queue + staffing math, macro library, escalation paths, QA + CSAT measurement, and a weekly support-health digest. Support is a retention lever, not a cost center.

Stand up and run a real customer-support function end to end. Pick the operating model, then route to helpdesk design/run, knowledge-base deflection, a live AI support agent, retention rescue, or a product-bug bridge, and assemble the pieces into one engagement plan. Human + AI-augmented; governed by grounded-only answers and no autonomous money movement.

Architect a self-serve knowledge base / help center that deflects tickets, mine ticket trends for gaps, design a job-framed taxonomy, author articles to archetype templates, stand up the instrumented deflection funnel, and optionally ground an AI answer bot that cites and escalates. The job is to prevent the next 100 tickets of this shape, and prove it.