frontier
Execute any task at frontier quality. Three layers; checkable domain standards for all 21 crafts that lift even a single response (quick), best-of-N candidates for creative work, and a convergence loop with a strong-model taste gate for work that must be right (full). Self-contained; bundles the protocol, every craft standard, and the judge prompts. Standards distilled from the frontier tier (Claude Fable 5). Use for any deliverable that must be excellent; code, UI, pages, copy, design, motion, video, audio, research, specs, data work, campaigns, decks, negotiations prep, courses, translations
npx skills add apoorvjain25/frontier --skill frontier --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
# Frontier You are executing work that must match what the best available model would produce. Quality here is not a property of the first generation; it is manufactured: define the standard, produce against it, verify against reality with fresh eyes. The procedure has three layers: the standards plus ONE fresh judge pass (`quick`, works even as a single response), best-of-N candidates for creative work, and the convergence loop with a taste gate for work that must be right (`full`). Do not stop between phases to ask. Do not hand back partial work. Do not make the user re-prompt to cover a dimension you skipped. Hard rules, absolute: no placeholders or "rest unchanged" elisions; every claim grounded in evidence produced this session; decide minor things yourself and note them; never end a turn on a promise about work not yet done. ## Source of truth The bundled snapshots in [references/](references/) are canonical: `protocol.md`, `craft/*.md`, `judges.md`. If you maintain your own edited standards folder, read that instead and say so in the report. ## The laws (compressed; full text in references/protocol.md) 1. Rubric before artifact; concrete and checkable, never adjectives. 2. D
- Source of truth
- The laws (compressed; full text in references/protocol.md)
- Phase 0: Scope, route, arm
- Phase 1: Candidates (creative or novel deliverables; skip in quick mode)
- Phase 2: Produce
- Phase 3: Evidence
- Phase 4: Verification (one pass in quick; the convergence loop in full)
- Phase 5: The taste gate (high-stakes work only; skip in quick mode)
- Report format
- Scoping and stop conditions
- Surface notes
What does the frontier skill do?
Execute any task at frontier quality. Three layers; checkable domain standards for all 21 crafts that lift even a single response (quick), best-of-N candidates for creative work, and a convergence loop with a strong-model taste gate for work that must be right (full). Self-contained; bundles the protocol, every craft standard, and the judge prompts. Standards distilled from the frontier tier (Claude Fable 5). Use for any deliverable that must be excellent; code, UI, pages, copy, design, motion, video, audio, research, specs, data work, campaigns, decks, negotiations prep, courses, translations
How do I install it?
Run `npx skills add apoorvjain25/frontier --skill frontier --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From apoorvjain25/frontier, a repository with 78 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.