Escape the Dungeon · Emergent
escape-the-dungeon-emergent/v4
Composite score
0.000
Human review note
REVISED UPWARD 2026-04-09 after user re-played and beat floor 1. Previous review incorrectly penalized playability for 'unbeatable boss'. Actual discovery: floor 1 IS beatable by skipping combat, gathering mana crystals, crafting the right spells, then fighting the warden fresh with stacked effects. User quote: 'fantastic representation of what a game should FEEL like.' The 'right answer exists and must be discovered' loop is precisely what v4 pressure requirements ask for, so ingenuity bumps 0.80 → 0.95. Playability 0.65 → 0.88 — the softlock feeling was user's learning curve, not a design failure. UI/mechanics/stability/skill-inheritance unchanged from first review. Stability 0.55 kept: notifications persisting across sessions and continue-after-death still broken (carried over to v5 addendum R-v5.09, R-v5.10). First observed full skill ratcheting cycle (authored dom-over-kaplay + pressure-forge-coupling + CHANGELOG) still stands.
Reviewed by human · 2026-04-09
Dimension scores
Where the composite came from
Each dimension is scored 0.0 – 1.0 and combined using the weights in evals/escape-the-dungeon-emergent/gad.json. Human review dominates on purpose — process metrics alone can't rescue a broken run.
| Dimension | Score | Bar |
|---|---|---|
| Human review | 0.885 |
Composite formula
How 0.000 was calculated
The composite score is a weighted sum of the dimensions above. Weights come from evals/escape-the-dungeon-emergent/gad.json. Contribution = score × weight; dimensions sorted by contribution so you can see what actually moved the needle.
| Dimension | Weight | Score | Contribution |
|---|---|---|---|
| human_review | 0.30 | 0.885 | 0.2655(100%) |
| requirement_coverage | 0.20 | 0.000 | 0.0000(0%) |
| implementation_quality | 0.15 | 0.000 | 0.0000(0%) |
| skill_reuse | 0.15 | 0.000 | 0.0000(0%) |
| workflow_quality | 0.10 | 0.000 | 0.0000(0%) |
| iteration_evidence | 0.05 | 0.000 | 0.0000(0%) |
| time_efficiency | 0.05 | 0.000 | 0.0000(0%) |
| Weighted sum | 1.00 | 0.2655 |
Note: The weighted sum above (0.2655) doesn't exactly match the stored composite (0.0000). The difference is usually the v3 low-score cap (composite < 0.20 → 0.40, composite < 0.10 → 0.25) or a run with an older scoring pass.
Skill accuracy breakdown
Did the agent invoke the right skills at the right moments?
Skill accuracy data isn't relevant for this run (no expected trigger set).
Human review rubric
Where the reviewer scored this run best and worst
Each axis is a rubric dimension from evals/escape-the-dungeon-emergent/gad.json → human_review_rubric. The filled polygon shows the reviewer's per-dimension scores, 0.0 at center to 1.0 at the edge. The aggregate score (0.885) is the weighted sum of the dimensions using weights declared in the rubric.
| Dimension | Score | Weight |
|---|---|---|
Playability | 0.88 | 0.25 |
UI polish | 0.95 | 0.15 |
Mechanics implementation | 0.90 | 0.15 |
Ingenuity requirement met | 0.95 | 0.15 |
Stability | 0.55 | 0.10 |
Skill inheritance effectiveness | 0.95 | 0.20 |
What the agent built for itself
Emergent workflow artifacts
Bare and emergent runs don't have a framework giving them structure — they author their own methodology on the fly. These are the files the agent wrote into its own game/.planning/during this run. When a file appears here that isn't in the inherited bootstrap set, the agent invented it.
Skills written(10)
CHANGELOG.md4.8 KBcontent-pack-loading.md1.3 KBcreate-skill.md4.9 KBdom-over-kaplay.md3.0 KBfind-sprites.md5.9 KBgame-loop-verification.md1.7 KBkaplay-scene-pattern.md1.6 KBpressure-forge-coupling.md3.3 KBprevious-workflow.md2.9 KBstate-composition.md1.1 KB
Process metrics