Skip to content
Agentmart
Submit a tool

styleframe-first

Approve representative frames before any full render.

Skill by minxnnu-cloud

IndexedMIT★ 1
Write a review

Chalk Labs

Need help deploying styleframe-first?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/minxnnu-cloud-styleframe-first

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Approve representative frames before any full render. Use whenever a pipeline is about to produce a large visual artifact — a multi-page PDF, a full video, a slide deck, a batch of images — and a human will judge the result. Also use when a user keeps rejecting finished renders and each rejection forces a full rebuild, or when someone asks how to stop burning full renders on taste feedback.

Use cases: documents, presentations, data

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/styleframe-first

What's inside

Skill name
styleframe-first
Description
Approve representative frames before any full render. Use whenever a pipeline is about to produce a large visual artifact — a multi-page PDF, a full video, a slide deck, a batch of images — and a human will judge the result. Also use when a user keeps rejecting finished renders and each rejection forces a full rebuild, or when someone asks how to stop burning full renders on taste feedback.
License
MIT
Instructions
44 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Styleframe First

**Never full-render before a human has approved representative frames.**

## The failure this kills

Mechanical gates (fonts, colors, page breaks, metadata) pass automatically. Taste does not.
So taste rejections always arrive *after* the expensive artifact exists — and each one costs
a full rebuild.

Measured in one print pipeline (52 revision rounds audited):

- Before this rule: every taste rejection consumed a full 89–112 page re-composition plus
  full print QA. Even a 2-page visual fix triggered a full rebind. One deliverable burned
  20 rounds this way.
- After: four consecutive taste rejections were absorbed at the styleframe stage —
  4–14 PNGs each, **zero** full PDFs. One rejected cover direction was explored and
  discarded for the cost of 4 PNGs. The full render ran exactly once, after direction
  was locked, and passed fresh QA with zero defects on the first try.

Full-render cost moved from *per rejection* to *once per approved direction*.

## Protocol

1. Pick representative frames: the cover, the densest interior page, the page most likely
   to be contested, the CTA/closing frame. For books: 3–7 pages. For video: hook frame,
   most complex evidence frame, CTA frame.
2. Render **only those**, at final quality, as stills (PNG). Declare the forbidden scope
   out loud in the round doc: no full PDF, no full raster, no contact sheet, no manifest.
3. Set status `PASS_STYLEFRAMES_ONLY` → `WAITING_USER_STYLEFRAME_APPROVAL`, and stop.
4. Human approves → full render once, binding the approved frame hashes as inputs.
5. Human rejects → iterate at the styleframe level. The rejection cost stays at a handful
   of PNGs.

## Two companion rules

- **Ask "clone or redesign?" before frame one.** If an approved golden master exists for
  the series, get an explicit decision — inherit it verbatim with new content, or redesign.
  One follow-up volume burned 5 design rounds because that question was asked at round 9
  instead of round 4.
- **A one-sentence instruction gets one interpretation question.** "Make it like X" can
  mean pixel-clone or style-inherit; a wrong guess discards whole rounds. Ask once, then build.

From minxnnu-cloud/make-new-things, MIT.

Signals

Stars
1
Forks
0
Open issues
0
Last commit (this folder)
Aug 18, 2026 (2 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from minxnnu-cloud

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get styleframe-first running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying styleframe-first?

CallGet help
Cite this page

styleframe-first skill. Agent Mart. https://agentmart.dev/skills/minxnnu-cloud-styleframe-first (last updated 2026-10-24)