JasonColapietro/suede-creator-skills

Suede Creator Skills

A 67-skill toolkit for Claude Code and Codex: take broad outcomes full send, orchestrate multi-agent teams and OpenAI Codex CLI worker fleets, run code review with an A-F ship grade, and design AI evals.

By Jason Colapietro / Suede Labs AI

What it is

A free, MIT-licensed, broadly reusable pack of 67 public skill folders for Claude Code and OpenAI Codex. Each skill is a skills/<name>/SKILL.md file the agent loads on demand.

• Full Send: turn max-effort, max-agent, spare-no-compute, and fix-everything intent into one authorized controller, every useful non-colliding lane, adversarial reconciliation, and concise proof (suede-full-send). House line: "Never end your allocation above zero." Dry joke, not a literal token promise.
• Agent orchestration: wire complex changes into coordinated agent lanes with WIP collision detection, RFC mode, feature-flag strategy, rollback trees, and a handoff checklist that won't close without evidence (suede-agent-teams).
• Codex worker fleets: the Suede Fable Fleet — a Claude orchestrator decomposes a high-volume job, writes self-contained briefs, spawns parallel OpenAI Codex CLI codex exec workers, and reviews every output against acceptance criteria before anything ships (suede-codex-fleet).
• Code review + A-F ship grade: deep findings plus a blunt letter verdict across 7 evidence-backed lanes, with instant-F triggers and grade caps for auth and payment surfaces (suede-code, suede-code-review, suede-code-grader, suede-ship-gate).
• AI evaluation: turn LLM, RAG, classifier, and agent surfaces into AI-SPEC artifa