Paste-ready. Every claim points at main.

Copy for the folder that runs 30+ live sites and 8 iOS apps from one terminal.

Paste any block here into a post or a README and it will hold. Every claim points at something you can open on main: a skill folder, or the commit that changed it. The long-form explainer lives in PROMO.md, and this page mirrors COPY.md, so you can diff the copy against its own source.

The repo, and the page it generates

For GitHub, where the pack gets one line

One-line description

Open-source AI skills for SEO, marketing, and business operations in Claude Code and Codex.

Full description

Suede Creator Skills gives founders, marketers, and small teams reusable workflows for SEO audits, AI search visibility, conversion copy, and business operations. Run them in Claude Code or OpenAI Codex, with instructions you can inspect and adapt to your business. Start with a page or business brief to get prioritized audit findings, positioning and campaign plans, or a map of operational bottlenecks. The pack also includes code review, app delivery, and creator-rights tools. Original work is MIT licensed; adapted components retain their upstream notices.

Component stack

Suede evolves while you use it. Capabilities load as inspectable SKILL.md folders instead of piling into one memory page that every prompt has to carry. An installed skill costs its name and its description. The body loads only when a request matches, so adding a skill makes the pack sharper instead of heavier.

GitHub Pages note

This site is the GitHub Pages surface generated from the public suede-creator-skills repo, not a separate marketing site. The page itself is the proof. It ships from the same history as the skills it describes: one author, and every change is a commit on main.

From arrival to two install commands

Hero

Suede Creator Skills for SEO and business growth.

Subhead

Suede Creator Skills gives founders, marketers, and small teams reusable workflows for SEO audits, AI search visibility, conversion copy, and business operations. Run them in Claude Code or OpenAI Codex, with instructions you can inspect and adapt to your business.

Proof line

Invoke suede-workflow-skills when the work crosses lanes, and name one skill when it does not. Routing is not guesswork: every skill description carries a NOT FOR line that hands the request to the sibling that should beat it.

Cracked Devs page, technical line

The Agentic Adderall Stack: four orchestration skills (ship, agent-teams, workflow-skills, recommend-next-action) indexed as a validator-pinned stacks entry in the catalog. One controller, disjoint file-ownership lanes, adversarial review inside the pass, and an evidence gate at the close. Membership drift between the catalog and the page fails CI. The Suede Thought Graph search caps light, standard, and deep at 55, 110, and 200 calls.

First install is two commands. After that, one.

Public installs and MCP

Install the skills from the public GitHub repo. The MCP server is optional: a dependency-free stdio server with 9 tools, 7 resources, and 5 prompts, and it ships no binaries and no telemetry.

Hero

Add the marketplace. Install the pack. The folder that runs Suede Labs is on your machine in about 30 seconds.

CTA options

Use suede-code to review my staged diff and give it an A-F ship grade. List the Suede skills that match this task and name the one that should lose. Run a copy audit on this page and return the exact rewrites. Build a QA checklist and stop at the first gate I cannot prove.

SEO/AEO/AI EO audit copy

Use Suede Copy when a page reads fine and still gets skipped. The audit scores lanes against evidence instead of offering opinions, and it hands back the exact line to replace the one that failed.

Audit promise

Find what blocks discovery and the next click, then get the exact line to replace it with.

Visibility grader copy

Suede Visibility Grader ends in a letter. It reads a public page the way a person and an answer engine both would, and a broken primary CTA caps the page at D no matter how well the rest reads.

One-line description

Grade a public page on findability, first-screen clarity, CTA pull, proof, and AI readability. Caps beat the average: no live inspection holds the page at C, and a broken primary CTA holds it at D.

CTA options

Grade this page A-F and tell me which cap decided the letter. Rewrite the first screen so the offer lands before the fold. Replace the CTA with one that names the result. Show me what an answer engine would quote from this page.

Code grade and agent-team copy

Use suede-code when a diff needs more than a polite review. It grades a staged diff A-F across seven lanes, from correctness and security through tests and deploy readiness, and the weakest lane caps the grade. A hardcoded secret, or an auth check you can walk around with a request param, locks it at F with the file and the line, and no other lane lifts it. Use suede-ci-gate when the verdict has to block the merge.

Agent-team loop copy

Split major work into lanes that own separate files and report to one controller. WIP collision detection keeps two lanes off the same file, rollback trees keep a bad lane from taking the branch with it, and the handoff checklist will not close without evidence.

Ship DAG copy

Use Suede Graph Flo XR when the outcome is broad, authorized, and worth maximum useful effort. It writes competing plans for the change, scores them, prunes, refutes the survivors, and builds only the winner. Light, standard, and deep cap the run at 55, 110, and 200 calls. It reads production for evidence and never deploys, and it does not expand permissions, spend authority, or model limits.

The evidence a release needs before it leaves your machine

Creator Rights Package Builder

Package a creator folder into a local rights and provenance bundle: credits, splits, license notes, and a report of what is missing. It runs offline, writes local files, and leaves an unknown rights fact marked unknown. It organizes evidence and does not clear rights.

Page hero

Create a local rights package before the work leaves your machine.

CTA options

Create a rights package for this folder and list what is missing. Open the skill folder and read the file before you trust it, since every skill is plain Markdown. Run the package script against a real release. Pair it with the release linter and close the gaps it names.

Release Metadata Linter

A release with missing splits or an unclear sample status becomes a rights problem after it ships, not before. The linter reads the folder and writes the gap list while the release is still yours to fix.

Page hero

Find release gaps before they become rights problems.

CTA options

Audit this release folder and rank the gaps by what blocks the release. Read the lint rules in the repo and disagree with them before you run the lint. Open the lint script and see exactly what it reads. Create a rights package once the gaps are closed.

Six marketing firms. Not one registered the property.

Jason hired and fired six marketing firms over Suede's life. Not one registered Suede's property in Google Search Console, and he did not even know the tool existed until he got close enough to the work to see what had been skipped. You learn how poorly things were done when you can finally get involved, and no one will ever care as much as the founder/operator who has to live with the result. Suede Creator Skills are those misses written down as methods, and they now run 30+ live sites and 8 iOS apps from one terminal with zero employees.

Posts you can check before you post them

Ready to paste. Every block is verified against the repo.

Suede Creator Skills: open-source agent skills for Claude Code and Codex. One of them grades a staged diff A-F across seven lanes, and the weakest lane caps the grade. One author, and every skill folder gets edited when the practice changes.

Docs: https://skills.suedeai.ai/skills/
I published Suede Creator Skills, the public SKILL.md pack for Codex and Claude Code. It is the same folder that runs Suede Labs, published as is.

One skill grades a staged diff A-F across seven lanes, and a hardcoded secret locks the whole thing at F with the file and the line. Another argued $448.31 back out of a real Amazon account, including a refund denied after the return window closed. The utility tools run offline, write local reports, and keep private files off the network.

Repo: https://github.com/JasonColapietro/suede-creator-skills
The label is a joke. The mechanism is a DAG.

New in Suede Creator Skills: the Agentic Adderall Stack, the pack's four orchestration skills (ship, agent-teams, workflow-skills, recommend-next-action) indexed as a validator-pinned stacks entry in the catalog. One controller, disjoint lanes, adversarial refute inside the pass, an evidence gate at the close.

Page: https://skills.suedeai.ai/cracked.html
I added a curated stack to Suede Creator Skills and made the marketing do engineering work.

Cost is on the label. The Suede Thought Graph search caps light, standard, and deep at 55, 110, and 200 calls, only the selected winner mutates, and production inspection is read-only.

The Agentic Adderall Stack groups the pack's four orchestration skills (ship, agent-teams, workflow-skills, recommend-next-action). Instant release is one continuous subagent-driven pass: the controller fans every disjoint lane out at once, adversarial review runs inside the pass, and the close is an evidence gate. Extended release is the scheduling counterpart: a scorer ranks the next move on goal fit, unblocking, evidence, urgency, and leverage, then emits it as a runnable prompt.

The grouping is a cross-cutting overlay in the catalog, not a seventh category. The specialty partition stays total, the validator fails CI if the stack's page and its catalog membership drift, and the MCP server scopes the stack per profile.

Page: https://skills.suedeai.ai/cracked.html

It runs offline. These are the limits, in plain words.

  • Public v1 runs offline and writes local files. It ships no binaries and no telemetry.
  • Generated reports are drafts until a creator or operator reviews them.
  • These skills organize evidence. They do not provide legal clearance.
  • The scripts skip hidden files, dependency folders, build outputs, caches, and secret-like files by default.
  • Unknown rights facts stay marked as unknown.

Proof is part of the release.

First install takes two commands. After the marketplace is added, install is one. Every skill is plain Markdown.

$ /plugin marketplace add JasonColapietro/suede-creator-skills → /plugin install suede-skills@suede All install paths