meta-skills

Decomposes a spec or architecture into buildable tasks with acceptance criteria, dependencies, and risk-first implementation order, for AI agents or human engineers. Produces `docs/forsvn/artifacts/meta/tasks.md`. Use once requirements are clear and you need an execution plan. Not for clarifying unclear requirements (use discover) or designing architecture (use architect-system). For code review after building, see review-work.

Designs technical blueprints — tech stack selection, database schema, API design, file structure, and deployment plan for a defined product or feature. Produces `architecture/system-architecture.md`. Not for unclear requirements (use discover) or task decomposition (use breakdown-tasks). For user journey mapping, see map-user-flow. For code quality after building, see review-work.

Designs the in-product interface from a validated map-user-flow — screen inventory, a reusable component system, DESIGN-token application, per-state visual treatments, accessibility floor, and a buildable handoff. Emits a portable SPEC, never rendered UI. Use after a user flow exists and before architect-system. Not for brand identity (create-brand), conversion landing pages (brief-landing-page), or technical architecture (architect-system).

Audits and refactors existing code for readability, maintainability, and dead code removal without changing behavior. Produces `docs/forsvn/artifacts/meta/records/[date]-cleanup-<slug>.md` and applies fixes in-place. Not for diagnosing business problems (use diagnose) or writing documentation (use write-docs). For writing missing docs after cleanup, see write-docs.

Audit + clean a developer's machine — dotfolders, caches, language toolchains, package globals — with per-target classification, risk surfacing (auth, processes, side effects), and explicit confirmation. Produces `docs/forsvn/artifacts/meta/records/machine-cleanup-*.md`. Not for cleaning code (use clean-code) or for triaging user files (those need human review).

Extracts repeated operational mechanics (SDK, API, file-system, or network logic copy-pasted across handlers or actions) into a shared service layer — produces a stepwise migration plan, then applies it caller-by-caller with verification at each step. Use when the same "how" recurs across 2+ callers. Not for behavioral cleanup or dead code (use clean-code) or designing a system from scratch (use architect-system). For a quality review after migrating, see review-work.

Maps multi-step in-product flows — screens, decisions, transitions, platform-native touchpoints (dock, menu bar, widgets, notifications, Live Activity), edge cases, and error states for a feature or user journey. Use when designing or auditing a feature spanning multiple screens, states, or platforms. Not for visual brand design (use create-brand), landing-page architecture (use brief-landing-page), or technical architecture (use architect-system).

Generate documentation from a codebase — READMEs, API references, setup guides, runbooks, architecture docs, ship logs, release notes (CHANGELOG + GitHub Release bodies). Not for specifying what to build (use discover), restructuring code (use clean-code), or task decomposition (use breakdown-tasks).

Produces production-ready briefs for app onboarding, App Store preview, and feature-demo videos from supplied UI screenshots. Defines the feature promise, cropped UI beats, interaction choreography, captions, motion rules, platform specs, and asset manifest. Use when screenshots exist and the goal is to show a product feature in action. Not for generic short-form social videos (use brief-shortform) or rendering the video (use produce-video).

Produce a graphic-design brief for a single visual asset — social posts, thumbnails, banners, OOH, OG cards, hero illustrations. Pulls brand tokens, generates concepts, writes platform-aware specs + image-gen prompt or designer-handoff spec. Does NOT render — rendering is downstream. Not for brand identity (use create-brand), whole-page redesigns (use brief-landing-page), or the copy that goes IN the asset (use write-copy).

Generates a campaign-grade brief for a conversion landing page or redesign — hypothesis, surface rhythm, section-by-section spec, asset slots, copy candidates, hand-off prompts, conversion + brand-voice critics. Output ready for Claude Design, Figma, a human designer, or a coding agent. Not for post-launch CRO from analytics (use evaluate-landing-page inside an eval-loop), non-conversion pages (blogs, docs), or a single visual asset (use brief-graphic).

Production-ready brief for short-form video — hook, shot list, on-screen text, audio plan, caption, CTA, aspect, length. Covers live-action + motion-graphic. 1 hero + 2 variants per run. Reads the per-platform research-shortform catalog. Not for static visual assets (use brief-graphic), organic social copy (use write-social), paid ad creative (use write-ad), or rendering (use produce-video).

Builds a brand identity system as up to five artifacts — BRAND.md (story, voice, archetype), DESIGN.md (design tokens), CREATIVE-DIRECTION.md (art direction), FRAME.md (frame direction for the camera), ASSETS.md (per-platform inventory). Runs Quick Brand (Route A) or full brand-system (Route B). Use to define or rebrand a product's identity. Not for marketing copy (use write-copy), user flows (use map-user-flow), campaign planning (use plan-campaign), or audience research (use research-icp).

Design a marketing mini-game / interactive — spin-the-wheel, quiz, scratch-card, playable ad, advent-calendar — for engagement, lead capture, or virality. Produces a game-design spec: the mechanic + win condition, the reward economics, the data-capture value exchange, and the conversion funnel. Honest by construction (no fake-win dark patterns). Not for building the game's code (that's an engineering task), nor a static ad/asset (use write-ad / brief-graphic / produce-asset).

Score a launched paid ad (Meta, Google, TikTok, LinkedIn) from real metrics — network-specific primary metric + diagnosis + fatigue/auction signals — inside an existing eval loop. One cycle per network. Not for loop setup (use run-pipeline), writing new creative (use write-ad), channel-mix retrospectives (use plan-campaign), or campaign-level scoring (use evaluate-campaign).

Score a produced visual asset (image / graphic / carousel frame) against its brief — render fidelity + brand-fit — inside an existing eval loop, using the re-ingested asset (the return-leg). One asset (or picked variant) per cycle, verdict + brief-fidelity diagnosis. Not for video (use evaluate-shortform), a landing page (use evaluate-landing-page), live-post engagement (use evaluate-content / evaluate-ad), generating the asset (use produce-asset), or scaffolding the loop (use run-pipeline).

Score a launched multi-channel campaign from real metrics inside an existing eval loop — verdict, per-channel breakdown, unit economics. Aggregate-only: one cycle = whole campaign, all channels. Requires an eval-loop workspace. Not for loop setup (use run-pipeline), re-planning (use plan-campaign), or per-asset eval — single ad uses evaluate-ad, post evaluate-content, page evaluate-landing-page, short-form evaluate-shortform.

Score a published organic post (text / image / carousel) from real metrics inside an existing eval loop — one primary platform per cycle, verdict + engagement-quality diagnosis. Not for short-form video (use evaluate-shortform), paid-ad performance (use evaluate-ad), writing next-cycle copy (use write-social), or scaffolding the loop (use run-pipeline).

Scores a launched landing page from real performance evidence inside an existing eval loop — writes a cycle eval snapshot and appends the loop results ledger. Use for post-launch CRO cycles backed by analytics, experiment results, recordings, form-funnel data, or qualified manual metric notes. Not for designing the next page brief or a redesign (use brief-landing-page), channel strategy (use plan-campaign), best-practice audits without evidence, or scaffolding the loop (use run-pipeline).

Score a sent outreach cycle (cold email / LinkedIn DM / sequence) from real reply data inside an existing eval loop — one channel + segment per cycle, verdict + reply-quality diagnosis with a deliverability + compliance gate. Not for writing the outreach (use write-outreach), organic-post performance (use evaluate-content), paid-ad performance (use evaluate-ad), or scaffolding the loop (use run-pipeline).

Score an SEO / AEO cycle from real ranking + visibility data inside an existing eval loop — one keyword cluster + surface (organic SERP or AI answers) per cycle, verdict + visibility-delta diagnosis with a lag-and-volatility gate that refuses short-window noise. Not for running the audit/fixes (use optimize-seo), tracking AI citations (use monitor-aeo), organic-post performance (use evaluate-content), or scaffolding the loop (use run-pipeline).

Produce a minute-by-minute run-of-show (cue sheet) for a live marketing event — webinar, launch event, AMA, demo day, product reveal, conference talk. Segments + timing + owners + technical cues + contingencies. Not for the campaign around the event (use plan-campaign), the launch-day social sequence (use the launch chain / publish-social), or a recorded video (use brief-shortform/produce-video).

Strip AI tells, inject brand voice, and compress existing text so it reads human-written — 15%+ word reduction with zero idea loss + optional detector-resistance pass. Use as a terminal polish pass on AI-drafted copy that sounds robotic. Not for writing new copy from scratch (use write-copy); for brand voice of record see create-brand.

Ingest a launch/campaign's real post-publish results for ONE channel, produce a structured what-worked/what-failed read against that channel's playbook pack, and feed a dated entry back into the pack so the next launch compounds. Closes the metrics loop (pillar 3). Not for scoring a single organic post in an eval loop (use evaluate-content), paid-ad results (use evaluate-ad), short-form video (use evaluate-shortform), or landing-page CRO (use evaluate-landing-page).

Monitors AI-search visibility across answer engines and generative-search surfaces — provider/query citation checks, Google AI Overview/GEO citation presence, Bing/IndexNow readiness, AI referral traffic, llms.txt status, competitor cited-domain share, and dated snapshot reports. Use when you need evidence of whether AI systems mention or cite a brand/site. Not for deciding SEO strategy or rewriting pages (use optimize-seo) and not for technical-audit remediation.

Audits and plans search visibility across six modes — technical SEO audit, AI/answer-engine optimization (AEO), programmatic SEO, competitor comparison pages, full SEO strategy, and app store optimization (ASO). Covers keyword research, on-page and technical fixes, link-building strategy, and structured data. Use to diagnose a traffic drop, plan search growth, or get found by AI search. Not for landing-page conversion brief work (use brief-landing-page) or writing the page copy (use write-copy).

Builds an integrated marketing plan — growth motion, messaging pillars, per-channel angles across all 9 channels, content calendar, budget allocation, and a phased go-to-market timeline. Use when you need a campaign strategy or GTM plan grounded in ICP research. Not for setting numeric funnel targets (use plan-funnel), search-channel strategy (use optimize-seo), or landing-page architecture (use brief-landing-page).

Polishes already-translated Vietnamese text so it reads natively in a target register (báo chí, semi-casual, bro, or pop-marketing) — fixes pronoun drift, missing particles, literal idioms, passive-voice calques, typography, and translationese. Post-translation polish only; does NOT translate from other languages. Use on translated Vietnamese copy that reads stiff or foreign. Not for tone work on English (use humanmaxxing) or writing new Vietnamese from scratch (use write-copy).

Previews, validates, and fixes Open Graph / Twitter share cards — how a link renders on Facebook, X, LinkedIn, and Discord/Slack. Runs the metaprev CLI to catch missing or broken og:image, relative URLs, wrong dimensions, oversized files, and weak titles, then applies fixes and re-verifies. Use before shipping a page or when a share preview looks broken. Not for page copy (use write-copy), search/AEO (use optimize-seo), og:image artwork (use produce-asset), or posting (use publish-social).

Turns a brief-graphic artifact into render-ready prompts + an asset manifest. Tool-agnostic by design — emits prompts tuned for your chosen image engine (Midjourney / DALL·E / Imagen / Figma / designer); the stack holds no API keys and runs no render engines. Use when a design brief exists and you need production-ready prompts. Not for the copy that goes in the asset (use write-copy), the brief itself (use brief-graphic), or publishing the rendered asset (use publish-social).

Produce an out-of-home (OOH) advertising brief + render-ready spec for billboards, transit, posters, and building wraps — built around the 3-second read, legibility at distance/speed, one message, and per-placement format/production specs. Emits a brief + render manifest (no pixels — an operator-connected engine renders). Not for digital/social ad creative (use write-ad / brief-graphic / produce-asset), nor the media buy/campaign (use plan-campaign).

Turn a brief-shortform or brief-app-preview into a multi-runtime export bundle — manifest, per-shot prompts, production-lane scaffolds, and a post (assemble · grade · subtitle) stage. Two modes: shortform (social/promo) and app-preview (screenshot-driven product demo). Tool-agnostic — emits scaffolds + prompts and routes a script to a production lane; does NOT invoke render engines. Not for writing the brief (use brief-shortform / brief-app-preview) or publishing (use publish-social).

Turn write-social copy into a publishing bundle — per-platform drafts plus scheduler-import files (Typefully / Buffer / Hootsuite / CSV). `--mode=publish` posts live behind a two-stage gate. Covers 9 platforms (X / LinkedIn / Instagram / YouTube / TikTok / Facebook / Bluesky / Threads / Reddit). Not for writing copy (use write-social), media generation (use produce-asset / produce-video), or landing-page placement.

Writes paid-ad copy across Meta, Google, TikTok, and LinkedIn — network-aware char-caps, policy floor, audience/intent framing, 7-dim rubric. Not for landing-page headlines (use write-copy), cold-outreach DMs (use write-outreach), organic social posts (use write-social), or channel-mix/budget strategy (use plan-campaign).

Writes and evaluates persuasive copy — headlines, hooks, CTAs, taglines, and landing-page sections — with per-line V/F/U rubric scoring, annotations, and ranked alternatives. Not for AI-sounding cleanup (humanmaxxing), search/AI-citation (optimize-seo), brand voice (create-brand), landing-page architecture (brief-landing-page), social posts (write-social), or paid-ad copy (write-ad).

The complete native launch-day copy bundle for ONE launch channel (Product Hunt, Reddit, any channel with a launch-channel pack): tagline/title, description, anchor narrative, notify + cross-post copy, metadata — pack-driven, with the pack's §4 hard guards enforced (PH no vote-ask; Reddit founder disclosure + 9:1). Single-channel; 5-dim critic. Not single social posts (write-social), the run-of-show (plan-campaign), gallery assets (brief-graphic), or post-launch reads (measure-results).

Writes and evaluates cold outreach — email, LinkedIn, Twitter/X, iMessage/SMS, platform proposals — with signal-based personalization, channel-specific craft, and rubric scoring. Handles first-touch compose and reply-to-inbound modes. Use to draft or critique an individual outbound message. Not for campaign orchestration or sequence design (use plan-campaign), landing-page or headline copy (use write-copy), paid-ad copy (use write-ad), or sourcing/list-building. Brand voice: see create-brand.

Platform-native social copy (hook variants + body + CTA + format spec) for tiktok, reels, shorts, x, linkedin. Single-platform per artifact. 5-dim rubric. Not for paid-ad copy (use write-ad), landing-page copy (use write-copy), short-form video briefs with storyboards (use brief-shortform), or long-form articles. Vietnamese polish: polish-vn. AI-tell cleanup: humanmaxxing.

Audit + groom `docs/forsvn/artifacts/` — classify every file (KEEP/STALE/ORPHAN/LEGACY/EPHEMERAL), surface references + risk. On `--apply` archives candidates behind per-category confirmation (moves to dated archive, never deletes). Not for cleaning code (use clean-code) or machine state (use clean-machine).

Runs a problem through multiple expert perspectives via debate (agents argue in rounds and converge) or poll (agents analyze independently, then aggregate by consensus). Use to pressure-test a decision or trade-off with no clear winner. Standalone, or invoked by another skill as a sub-routine. Not for implementation (use architect-system) or code verification (use review-work).