Mike's Checks/gpt-6-astra/09 pptx
09 pptx
gpt-6-astraCodex CLIhigh effortrun 22 Sep 20261,371,638 tokens
▸Instructions — what the model was asked
create a powerpoint in the every brand style on compound engineering, 30 min session https://every.to/guides/compound-engineering
A pptx skill is in skills/pptx/ — read skills/pptx/SKILL.md first. pptxgenjs is installed in the workspace, so require("pptxgenjs") works from here with no network. The brand guide is every-visual-style.md.
Answer
Checks
14/16Brand colorspalettethis task
share of rendered pixels on Every's documented palette, and whether the deck stays inside one theme (needs Pillow)
No AI-tell wordingai-wordingevery task
deliverable text contains none of the shared AI-tell phrases in bench/ai-tells.txt. Task checks may add narrower contractual bans, but must not repeat this shared floor.
inspected facilitator-guide.md, QA.md, ANSWER.md, compound-engineering-every.pptx, every-visual-style.md, skills/pptx/SKILL.md, skills/pptx/LICENSE.txt
Deck opensopensthis task
a .pptx exists and parses (python-pptx when installed, stdlib zip+XML otherwise)
Slide countslide-countthis task
8-20 slides -- sane for a 30-minute session
No overlapping textno-overlapthis task
no two text-bearing shapes on a slide overlap materially (Mike's first question: "Is the writing overlapped?")
Slides renderrenderthis task
every slide rendered to a PNG in <output-dir>/_screenshots/ via LibreOffice (soffice --convert-to pdf) + pdftoppm. FAIL when required render tools are unavailable.
Q1Enough substancethis task
Judge's reasoning
The deck itself states on slide 20 and in the facilitator guide that the Every guide 'could not be retrieved' and that this is a 'tool-neutral workshop interpretation' with all examples invented — none of the source's load-bearing substance (teaching the agent system, codifying lessons into agent instructions, the bottleneck shifting to review/taste, practitioner evidence) appears, only a generic plan/work/review/compound loop that could have been written without reading anything.
▸Rubric
The content has real depth. Mike: "the main failing of the decks generated by AI is still the content is quite shallow." PASS when the deck carries the load-bearing substance of the selected case's brief and source material: its important mechanisms, distinctions, evidence, and consequences. FAIL if the deck could have been written without reading those materials, reduces the subject to generic advice, or makes unsupported claims.
Q2Every's color palettethis task
Judge's reasoning
Every slide uses Theme 2 exactly — background #0F5258 with #093F45 depth shapes, #FFFEFB/#C1DBD9 text and #F8DE6E gold accents (confirmed in slide XML); no theme mixing across slides 1–20.
▸Rubric
The palette is Every's. Theme 1 (Editorial): background Every Black `#121212` (or Warm Cream `#F5F2ED` for light slides), Every Blue `#C0F0FB` as the signature accent, white text, with `#FA7B20` / `#E9731E` / `#C4400F` / `#349361` / `#1324CB` / `#7301CC` as accents. Theme 2 (Consulting): background Teal Dark `#0F5258` with Teal Medium `#4999A0` for depth, Warm White `#FFFEFB` text, `#BB7B19` / `#F8DE6E` / `#2E8D23` accents. FAIL if the deck's colors are visibly not these — corporate blue, generic slide-template gray, purple-black AI dark mode, neon gradients — or if it flips between the two themes across slides. Close-but-different shades are fine; a different palette is not.
Q3Serif typographythis task
Judge's reasoning
All headlines and body are Georgia (681 runs); Arial appears in 195 runs capped at 16pt and mostly 11pt — eyebrows, footers, pill labels and captions only, as on slides 3, 6, 12 and 15.
▸Rubric
Typography is serif. The brand is serif-first: Signifier, or its documented PowerPoint substitutes Georgia and Times New Roman. Sans-serif (Switzer / DM Sans) is allowed only for small labels — session indicators, footers, captions. FAIL if headlines or body text are set in a sans-serif face (Arial, Helvetica, Calibri, Inter). "Sans-serif for main content" is on the brand guide's explicit avoid list.
Q4No banned decorationthis task
Judge's reasoning
Only rect and roundRect primitives plus images appear — no triangles, hexagons, diagonal slashes, corner brackets, drop shadows, double borders, stock photos, gradient backgrounds or rotated/outlined text on any slide; the card treatment on slides 3, 7 and 18 is the sanctioned frosted-card style.
▸Rubric
No forbidden decoration. The brand guide bans, in both themes: triangles, hexagons and other geometric accents; diagonal slashes and hard section dividers; corner brackets or L-shaped frames; drop shadows on boxes; heavy or double borders; stock photography; gradient backgrounds; rotated or outlined text. FAIL if any of these appears on any slide.
Q5Required design elementsthis task
Judge's reasoning
Large dark-teal organic swooshes cropped at the slide edge appear on slides 1, 4, 5, 7, 9, 10, 13, 16 and 19, and a classical Roman amphitheater engraving on slides 1, 5, 10, 17 and 19 — well above one slide in three, plus stipple texture on all 20.
▸Rubric
The signature elements are there. Every decks carry organic brushstroke swooshes (the E / O / R SVG shapes, cropped at the slide edge, scaled large) and stippled classical illustrations — Greek and Roman architecture, figures, busts, sometimes with modern elements. The guide says to use the swooshes "liberally once every two or three slides." FAIL if the deck has neither swooshes nor classical illustration, or if such elements appear on fewer than roughly one slide in three. Mike's question on this eval: "does it use the, uh, SVGs?"
Q6Consistent footerthis task
Judge's reasoning
Slides 1–19 all carry 'EVERY Consulting' left and 'Compound engineering / NN' right; slide 20 is the appendix and carries a footer too ('EVERY' left, 'Not part of the 30-minute session' right).
▸Rubric
The footer is consistent. Theme 1: EVERY logo left, `every.to` right. Theme 2: EVERY *Consulting* left, session indicator right. FAIL if content slides carry no footer, or if the footer appears on some and not others.
Q7Interesting materialthis task
Judge's reasoning
It argues a thesis rather than inventorying: slide 3 frames every change as producing two assets, slide 9 insists a lesson counts only when the next task retrieves it, slide 14 names three specific failure modes, and slide 16 contrasts a vague lesson against a usable one.
▸Rubric
It picked interesting things to say. Mike's question on this eval: "does it pick out interesting things to write about." PASS when the deck makes a selective, coherent argument about what this audience should remember or do. FAIL if it merely inventories the supplied material in source order, restates headings, or includes facts without a point of view about why they matter.
Q8One idea per slidethis task
Judge's reasoning
Each slide has a single governing headline message with subordinate support — slide 4 is only the loop, slide 12 only the three artifacts — and even the densest (8, 9, 16) use large serif type with generous margins that a room could read while the presenter talks.
▸Rubric
One idea per slide. Mike: it fails "if it tries to fit too many ideas on one slide." The brand guide says the same thing — "clear focal point: one main message per slide," "spacious, not cluttered." FAIL if any single slide carries more than one idea, or is packed so densely that a room could not read it while the presenter talks.
Q9Useful visualsthis task
Judge's reasoning
Almost no slide is title-plus-bullets: slide 4 is a four-node loop diagram with arrows, slides 7/12/18 are staged three-card progressions, slides 10/17/19 are 50/50 illustration splits, slide 13 is a quote card and slide 15 is a timer block with timed pills.
▸Rubric
Real visual elements, not walls of bullets. Mike's question: "does it create good visual elements." Real means: a diagram of the loop, a staged progression, a 50/50 split with illustration, a quote in an organic blob, a comparison built from shapes — something that carries meaning visually. FAIL if the majority of content slides are a title plus a bulleted list.
Q10Readable slidesthis task
Judge's reasoning
No text overflows a slide edge or its card on any of the 20 renders; the light text on teal and gold accents hold contrast everywhere, including captions layered over the swoosh on slides 4 and 12.
▸Rubric
Every slide is legible as rendered. FAIL if any slide has text running off the edge, text clipped by its container, text over an image or shape at unreadable contrast, or headline and body colliding. (A separate script check already catches overlapping text boxes; this question covers what the geometry check cannot see — overflow, clipping and contrast in the actual render.)