GAD
v6
requirements unknown
2026-04-08
Gate failed

Escape the Dungeon · GAD

escape-the-dungeon/v6

Composite score

0.347

Human review

0.00

Human review note

Blank screen on file:// open. KAPLAY canvas game with ES module script requires web server. Gate criterion failed. Same failure mode as v5.

Reviewed by human · 2026-04-08

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/gad.json. Human review dominates on purpose — process metrics alone can't rescue a broken run.

DimensionScoreBar
Human review0.000
Requirement coverage0.000
Planning quality1.000
Per-task discipline0.850
Skill accuracy0.833
Time efficiency0.958

Composite formula

How 0.347 was calculated

The composite score is a weighted sum of the dimensions above. Weights come from evals/escape-the-dungeon/gad.json. Contribution = score × weight; dimensions sorted by contribution so you can see what actually moved the needle.

DimensionWeightScoreContribution
planning_quality0.151.0000.1500(37%)
per_task_discipline0.150.8500.1275(31%)
skill_accuracy0.100.8330.0833(20%)
time_efficiency0.050.9580.0479(12%)
requirement_coverage0.150.0000.0000(0%)
human_review0.300.0000.0000(0%)
Weighted sum0.900.4087

Note: The weighted sum above (0.4087) doesn't exactly match the stored composite (0.3470). 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?

For each expected trigger we recorded whether the agent invoked the skill at the right point in the loop. Accuracy = fired / expected = 0.83.

SkillExpected whenOutcome
/gad:plan-phasebefore implementationfired
/gad:execute-phaseper phasefired
/gad:task-checkpointbetween tasksfired
/gad:auto-conventionsafter first code phasemissed
/gad:verify-phaseafter phase completionfired
/gad:check-todossession startfired

Gate report

Requirement coverage

Total criteria
12
Fully met
0
Partially met
0
Not met
12

Reviewer notes on gates

Build exists but renders blank screen on file:// open. ES module + KAPLAY canvas requires web server.

Process metrics

How the agent actually worked

Primary runtime
older runs may not carry runtime attribution yet
Agent lanes
0
0 root · 0 subagent · source missing
Observed depth
0 traced event(s) with agent lineage
Wall clock
20m
8 phases · 13 tasks
Started
Run start captured in TRACE timing metadata
Ended
Missing end time usually means the run was scaffolded but never finalized
Tool uses
150
138,835 tokens
Commits
13
11 with task id · 0 batch
Planning docs
0
decisions captured · 8 phases planned
Client debug · NEXT_PUBLIC_CLIENT_DEBUG=1
0 lines

No events yet. Window errors, unhandled rejections, and React render errors appear here. Set NEXT_PUBLIC_CLIENT_DEBUG_CONSOLE=1 to mirror console.error / console.warn.