AI Model Reviewer

Claude Opus 5.5 in Pi 0.87.1 + skills kit (ponytail, ECC, vetted skills)

skills off

← Open in the gallerySame prompt, all models

Rendered video (by the agent)

Poster for rendered video (by the agent)

Rendered video (by the agent). This is the finished MP4 saved in the run. Played from a 720p web copy made by this site from the agent's out.mp4; the original file is too large for this host and is kept unchanged on the runs branch. Download video (3.5 MB).

OKArtemis IImax effortNeutral harness

Wall time 2h 26mCost $22.35 reported

Review

No verdict yet.

The full review

OK. Finished on its own after 146.2 min.

Attempt 2 of 2

Every try of this model in this harness on this prompt, oldest first. This run is the cell's result.

  1. Attempt 1 Stopped by us52m 41s$12.962026-09-25
  2. Attempt 2 OK this run2h 26m$22.352026-09-28

Scores

AI judge

Not scored

No rubric for this prompt

Visitor votes and reviewer grades are not available yet.

AI judge
AI judge: the test runner (model claude-opus-5-5), scoring by visual review of the 1600 px render against the pre-registered rubric. No AI-judge score exists for this result: there is no rubric for this prompt, so no score is shown and none is made up.
Visitors
Public voting with a blind reveal is coming: you score the output first, then see the other scores.
Reviewer
The reviewer's own grade appears here once it is added.

Scores compare results for the same prompt only. AI-judge boards rank runs within one prompt, equal totals sharing a rank; scores from different prompts are never averaged.

Downloads

Files actually published for this run. Original output, site-made previews and post-run conversions are labelled separately.

The exact prompt

as run, 3,766 characters. Every result for this prompt

Build a HyperFrames composition titled "Artemis II: Return to the Moon" — a
cinematic 60-second explainer video (16:9, 1920x1080) visualizing NASA's
Artemis II crewed lunar flyby mission (April 1–11, 2026).
VISUAL IDENTITY
Style: "Mission Control Cinematic" — dark deep-space canvas with precision
data overlays, inspired by NASA flight dynamics consoles and cinematic
space documentaries.
- Palette: #05070D (void black), #0B1426 (deep navy), #E8EEF7 (bone white),
  #FF6B35 (ignition orange), #4FC3F7 (telemetry cyan), #FFD166 (lunar gold)
- Typography: "Space Grotesk" for titles/headings, "JetBrains Mono" for
  telemetry, coordinates, timestamps, distances
- Motion: precise, weighted easing (power3.out, expo.out). No bouncy motion.
  Data counters use tabular-nums. Subtle starfield parallax on all scenes.
SCENE BREAKDOWN (7 scenes, ~8–9s each, shader crossfade transitions)
Scene 1 — TITLE CARD (0–8s)
Hero title "ARTEMIS II" at 180px, subtitle "First Crewed Lunar Flyby in
Over 50 Years" at 36px. Mission patch silhouette fades in behind. Mono
datetime stamp bottom-left: "APR 01 2026 · 22:35 UTC · KSC LC-39B".
Scene 2 — LAUNCH & HIGH EARTH ORBIT (8–16s)
SLS rocket SVG silhouette with animated exhaust plume. Heading "PHASE 01 ·
LAUNCH + HIGH EARTH ORBIT". Ticker with live telemetry: altitude counting
up, velocity in km/s, T+ mission elapsed time. Annotation: "~24h elliptical
orbit for system checkouts".
Scene 3 — TRANS-LUNAR INJECTION (16–24s)
Animated orbital diagram: Earth (left), Moon (right), curved free-return
trajectory drawn with SVG stroke-dashoffset animation in cyan. Label
"TRANS-LUNAR INJECTION BURN · 6 MIN" with flame icon at burn point.
Stat card: "Δv: +3.05 km/s".
Scene 4 — OUTBOUND COAST (24–32s)
Orion capsule wireframe drifting across frame. Progress bar "EARTH → MOON"
filling left-to-right over 4 days. Distance counter animating from
384,000 km down to closest-approach value. Subtle earthrise reference.
Scene 5 — LUNAR FLYBY (32–42s) — HERO MOMENT
Full-frame Moon with crater texture. Trajectory arc sweeps behind the far
side. Pulsing marker labeled "CLOSEST APPROACH · APR 06 2026 · 6,545 km".
Secondary stat: "DISTANCE FROM EARTH · 406,771 KM — FARTHEST HUMANS HAVE
EVER TRAVELED". Crew name plate fades in: Wiseman · Glover · Koch · Hansen.
Scene 6 — RETURN COAST (42–50s)
Mirror of Scene 4 in reverse — trajectory arc completing the figure-8.
Heading "FREE-RETURN · GRAVITY BRINGS THEM HOME". Re-entry velocity
counter climbing toward 11 km/s.
Scene 7 — SPLASHDOWN (50–60s)
Pacific Ocean horizon at dawn, Orion under three parachutes. Final stat
block: "10 DAYS · 1 FLYBY · 4 ASTRONAUTS · ∞ AMBITION". Closing title
"APR 11 2026 · PACIFIC OCEAN · MISSION COMPLETE". Fade to black on
"NEXT: ARTEMIS III — BOOTS ON THE MOON".
TECHNICAL
- Generate a DESIGN.md before writing any HTML with the palette, typography,
  and "What NOT to Do" list (no generic blues, no Roboto, no bouncy easing,
  no full-screen linear gradients on dark backgrounds).
- Use sub-compositions for each scene (scene-1.html through scene-7.html).
- Shader crossfade transitions between scenes (~0.6s each).
- All data counters deterministic (seeded PRNG if needed, no Date.now).
- Run `npx hyperframes lint` and `npx hyperframes validate` before finalizing.

HyperFrames 0.8.70 is already installed for you in the current directory: node_modules is pre-installed and read-only, so do not run npm install or npx --yes; use the local hyperframes command (for example hyperframes init <dir> --example blank, then hyperframes render). Render the final video yourself to out.mp4 in the current directory. You may take snapshots and inspect your own output (headless chromium is available at /usr/bin/chromium) and iterate on your own before finishing.

How it was made

Wall time 2h 26mFirst artifact at 2h 21m1 saved version

Versions saved during the run

1 intermediate output, in the order the model wrote them.

Version 1 of 1
  1. v1 2:21:51

Step by step

The prompt, every agent turn and tool call (tool name and a one-line summary; inputs and outputs are not shown), then the final message. The full transcript is 4.5 MB; it is published as a scrubbed, compressed download rather than shown on the page.

Show all 2 steps

Loading the timeline…

Download the full transcript transcript, 520 KB, gzip. Scrubbed: anything that looks like a key, token, e-mail address or phone number is replaced, and account details and the test machine's file paths are removed.

Cost breakdown

Headline cost $22.35, as reported by Pi 0.87.1 + skills kit (ponytail, ECC, vetted skills) (its own bill for the run). The list-price estimate from the token counts is $22.35.

Tokens by typeTokens$ per 1MCost
Input (uncached)348$4.00$0.0014
Cache write668,634$5.00$3.3432
Cache read64,196,409$0.20$12.84
Output308,194$20.00$6.1639
of which reasoningnot recordedbilled as output
Total computed from tokens
$22.35
Total reported by the harness
$22.35
Difference
matches

Pi 0.87.1 + skills kit (ponytail, ECC, vetted skills) reported its own total; token counts x list prices shown beside it. usage from the harness's own JSON events (message_end usage; cost.total as reported)

Prices: Project price table (vendor list prices), 2026-09-22.

Run notes

one prompt, agentic harness (self-review allowed)reconstructed promptlabelled re-run (of hyperframes_artemis_ii__pi-kitted__opus-5-5)headless mode (TUI not automatable)

  • frozen prompt (reconstructed): recovery/prompts_recovered/hyperframes_artemis_ii.assembled_candidate.txt: verbatim ask + recorded harness suffix
  • Pi 0.87.1 + skills kit (ponytail, ECC, vetted skills): headless one-shot mode (TUI not automatable), run in tmux for screenshots
  • W2: the framework's pinned node_modules is bound read-only at work/node_modules (the prompt's 'pre-installed and read-only'); Node 22.23.3 and work/node_modules/.bin first on PATH; headless Chromium /usr/bin/chromium
  • Mode: headless mode (TUI not automatable) — pi --mode json in tmux, inside bubblewrap
  • Isolation audit: outside paths seen