/suede-workflow-skills · Public umbrella skill

One install. The whole discipline.

Install one public GitHub skill to route the agent across all 71 Suede skills: Suede Full Send for broad authorized outcomes, Johnny Suede Write, Johnny Suede Design, Suedify, coordinated agent teams, Codex CLI worker fleets, full code review, A-F code grading, CI ship-gates, AI evals, Instagram growth, a scored next-action recommender, anti-slop copywriting, visual QA, visibility and CTA grading, schema-level Suede SEO/AEO/GEO/AI EO, site polish, launch packaging, MCP QA, a site-to-iOS app conversion workflow, a one-prompt Android app factory, plus music/IP tools, artist campaign tools, and creator utilities.

Install the public skill View skill folder

"load the whole Suede workflow" /suede-workflow-skills Ask in plain words. The router reads the request and loads this lane; MCP agents find it with list_suede_skills.

$ /suede-workflow-skills 28 lanes loaded · router listening "grade this diff" -> /suede-code PACK LOADED · ASK IN PLAIN WORDS

Public install command

Claude Code — add the marketplace and install the pack:

/plugin marketplace add JasonColapietro/suede-creator-skills
/plugin install suede-skills@suede

Codex — installs from GitHub as a standard Codex skill folder:

python3 ~/.codex/skills/.system/skill-installer/scripts/install-skill-from-github.py \
  --repo JasonColapietro/suede-creator-skills \
  --path skills/suede-workflow-skills

Restart Codex after installing the skill.

What it loads

Suede Full Send for broad, authorized outcome-bound work: one controller, useful non-colliding lanes, consequential reasoning, adversarial review, proof, and a concise final brief.
Portable preferences so agents can link to context instead of dragging one bloated prompt from job to job.
Johnny Suede Write for one-name copy, company voice, product and mobile copy, SEO/AEO/GEO/AI EO, CTAs, launch copy, and anti-slop line editing.
Johnny Suede Design for one-name Suedify, mobile and product surfaces, product screenshots, GitHub Pages polish, design-system QA, visual QA, UI polish, responsive checks, visibility grading, implementation handoff, company voice, and copy.
Suedify for reference-site restyling through Johnny Suede Design, with reference analysis, screenshot checks, and token distillation.
CI ship-gates with stack and lockfile detection, path-aware builds, one required check, and branch protection that gates the merge.
Suede Design for product-aware UI direction, design-system artifacts, and visual QA.
Suede Copy for public copy, anti-slop line editing, SEO/AEO/GEO/AI EO metadata, docs, CTAs, and evidence boundaries.
Suede SEO Audit for metadata, schema, search intent, answer intent, GEO/AI answer readability, links, and trust.
Suede Visibility Grader for A-F website visibility, CTA pull, proof, AI readability, rendered design signal, evidence, grade caps, and Google and Gemini result-surface checks.
Suede Site Alchemy for page structure, CTA routing, and site polish.
Suede Code Grader for direct A-F ship-risk grades with explanations and required upgrades.
Project grading for design, copy, SEO/AEO/GEO/AI EO, code, launch readiness, proof, missing gates, and exact next fixes before handoff.
Suede Code Review for bugs, regressions, security, tests, published-statement drift, and fix briefs.
Suede Agent Teams for scout, safe parallel build, shared context, adversarial review, consensus review, WIP protection, quality/eval gates, release lock, root-cause-aware recovery, and evidence handoff loops on important work.
Suede Fable Fleet for reviewed Codex CLI worker volume, with up to three genuinely different fixes and an early stop when the same root cause repeats.
Launch packaging and MCP QA skills for public release proof, install paths, structured discovery, audit scaffolds, QA checklists, MCP tools, and docs alignment.
Artist campaign skills for eras, song worlds, hooks, stunts, fan rituals, visualizers, merch objects, setlists, catalog revival, identity, collabs, campaign boxes, and sync review packaging.
Creator utility skills for linting, local rights packaging, rights audits, provenance, licensing, royalty routing, split notes, metadata checks, and evidence tables when the project needs them.

How to explain it

Suede Workflow Skills make an AI agent more consistent across product work. Instead of rebuilding the same prompt stack, the agent gets portable preferences, sharper pages, cleaner docs, stronger QA, public-safe Suede language, and a GitHub Pages bar that looks like a real product.

Use MCP only when structure helps: discovery, install guidance, SEO/AEO/GEO/AI EO audit scaffolds, GitHub Pages QA, or repeatable checklists.

Cue Suede

Feedback can happen mid-workflow or at the end. Say Cue Suede to bring up choices: change something, preserve what worked so the agent can mimic it later, or keep as-is by saying nothing. That is how the skill evolves around your preferences without turning into one huge memory page. Final handoffs should explain the result twice: a simple non-coder version first, then the usual breakdown.

Visibility loop

Use the visibility grader when a public page needs a blunt read before promotion. It grades findability, first-screen clarity, CTA pull, proof and trust, AI readability, schema depth, and design signal. It can also check whether the page is prepared for Google, Gemini, and AI result surfaces without mislabeled screenshot receipts.

Read visibility grader docs

Agent grouping loops

For major work, split lanes by ownership: scout and constraints, safe parallel build, shared evidence, adversarial review, consensus review, design visibility review, A-F code grading, WIP protection, quality/eval gates, release lock, recovery replay, and evidence handoff. Use parallel lanes only when write surfaces do not collide. Recovery replay allows up to three genuinely different fixes per failing check and stops early when the same root cause repeats, surfacing it to the user instead of retrying blind.

Best prompts

Use $suede-workflow-skills to improve this public page, tighten the copy, check SEO/AEO/AI EO, and run QA.

Use $suede-workflow-skills to review this repo and docs before I share the install instructions publicly.

Use $suede-workflow-skills to make this GitHub Pages site feel like a real product surface, not a dumped README.

Use $suede-full-send to finish this broad, authorized release outcome with useful parallel lanes, adversarial review, and proof.

Use $suedify with this reference URL and target URL, then verify the result on desktop and mobile.

Use $suede-visibility-grader to grade this public page A-F for findability, first-screen clarity, CTA pull, proof, AI readability, and design signal.

Use $suede-code-grader to grade this code change A-F, explain why, and list the required upgrades before ship.

Use $suede-agent-teams to recover this failing check with a failure-reason ledger. Allow up to three genuinely different fixes, but stop early when the same root cause repeats.

All 71 public skill folders

suede-full-sendRoutes broad, authorized outcome-bound work to one controller, useful non-colliding lanes, adversarial review, and proof. suede-workflow-skillsPublic umbrella entry point for the full Suede workflow pack. johnny-suede-writeFull writing mode for copy, company voice, product and mobile copy, SEO/AEO/AI EO, CTAs, launch copy, variants, and anti-slop editing. johnny-suede-designFull design mode for Suedify, mobile and product surfaces, UI polish, visual QA, visibility grading, company voice, and copy. suede-designDesign direction, responsive QA, product UI polish, and design-system checks. suede-copyConversion copy with headline formulas, persuasion frameworks, A/B variants, an anti-slop gate, and a copy score. suede-deslopStrips AI writing patterns from prose before it ships: merged kill list, structure fixes, active voice, varied rhythm, and a 50-point score gate. suede-codeUnified code review and A-F grading for correctness, security, data/state, deploy readiness, and ship risk. suede-code-reviewDeep diff review for behavior, regressions, security, tests, release gaps, and fix briefs. suede-code-graderA blunt A-F ship verdict across seven evidence-backed lanes, with instant-F triggers and surface grade caps. suede-agent-teamsScout, safe parallel build, adversarial review, consensus review, WIP protection, release lock, recovery, and evidence handoff loops. suede-codex-fleetSuede Fable Fleet: Claude decomposes, briefs, and reviews; parallel OpenAI Codex CLI workers generate the volume. suede-ai-evalAI-SPECs, failure-mode rubrics, eval cases, acceptance gates, and coverage audits for AI-powered product surfaces. suede-recommend-next-actionScores candidate moves on goal fit, unblocking, evidence, urgency, and leverage, then hands back one recommendation as a short runnable prompt. suede-ci-gateAny-repo CI gate that blocks a merge when required checks fail. suede-seo-auditSearch intent, answer intent, metadata, schema, links, trust, and AI summary-safe proof checks. suede-visibility-graderA-F grading for findability, first-screen clarity, CTA pull, proof, AI readability, rendered design signal, evidence, grade caps, and Google and Gemini result surfaces. suede-site-alchemyLanding page, campaign page, CTA routing, conversion, and page polish workflow. suede-launch-packagingPublic releases, proof links, install commands, QA, and handoff notes. suede-mcp-qaMCP tools, prompts, resources, catalog output, and docs alignment. suede-campaign-in-a-boxArtist campaign workflow for rollout phases, copy, content calendar, fan actions, page sections, and next moves. suede-sync-packagingPrepares sync-style review packages without placement promises, clearance claims, outreach claims, or a Suede promo CTA. suede-release-linterRelease metadata, files, credits, splits, sample status, provenance, and readiness blockers. suede-rights-passportTransfer package with provenance, credits, split notes, license notes, intake JSON, and missing-info reports. suede-rights-auditOwnership, contributor, split, sample, license, and intake gap audit. site-to-ios-appURL audit, App Store 4.2 wrapper-risk checks, Capacitor or native-shell strategy, and iOS build scaffolding for turning a site into an app. android-app-factoryKeyword to Play Store: Kotlin + Jetpack Compose scaffolding, Play Billing, ASO listing copy, Data Safety disclosure, and a signed release.

Evidence boundary

The skill gives agents a repeatable workflow. It does not prove legal clearance, invent metrics, guarantee conversion, copy private assets, or replace human review for rights, payments, or release decisions.

Pair it with artist skills

Use Suede Creator Skills when the work needs an era, song world, hook map, campaign box, visualizer direction, merch object, collab angle, or creator utility report.

Open skill catalog