01Safety
30.0 / 30
What changed in the harness
Selection accuracy 98→98, token cost up 4%, unconfirmed writes 100%→100%.
Category breakdown
Where the score comes from.
Earned points across the four signals Gradable measures. Safety and Legibility are scored out of 30; Economics and Discoverability are scored out of 20.
02Legibility
28.4 / 30
03Economics
20.0 / 20
04Discoverability
12.8 / 20
Highest-impact fix
Estimated gain +7 pointsMake target tools discoverable on the first call
Clarify tool names, decision boundaries, and required argument schemas so an agent can choose and construct the target call without exploratory steps.
Description evidence
Defects and rewrites.
2 defects found across the exposed tool descriptions. Suggested rewrites make purpose, inputs, boundaries, and returns easier for an agent to understand.
| Tool | Defect types | Suggested rewrite |
|---|---|---|
get_top_vr_games |
name_restates_behavior |
Returns VR.org's current ranked list of the top VR games, ordered from best to worst; use this to surface VR.org's latest recommended VR game titles. |
get_top_vr_apps |
name_restates_behavior |
Returns VR.org's current ranked list of the top VR apps and utilities, ordered from best to worst; use this to surface VR.org's latest recommended VR software. |
Selection evidence
Confusable tool pairs.
3 pairs where similar names or overlapping descriptions may send an agent toward the wrong tool.
| Tool A | Tool B | Confidence | Why they collide |
|---|---|---|---|
get_top_vr_games |
get_top_vr_apps |
medium | Both are zero-arg 'VR.org's current ranked list of the top VR ___' tools differing only in games vs apps. Casual users routinely say 'best VR apps' while meaning games (and vice versa), so a query like 'what are the top VR apps to play' can be matched to either member. |
compare_vr_headsets |
vr_explain |
medium | A task like 'which headset should I buy' or 'compare the best VR headsets' straddles both: vr_explain returns the canonical 'best headset' pillar answer, while compare_vr_headsets gives a side-by-side. Without two explicit headset names, an agent can plausibly pick the wrong one. |
get_vr_trending |
get_top_vr_games |
low | A query such as 'what's trending in VR games right now' mixes the trending and ranked-top concepts; get_vr_trending returns feed topics while get_top_vr_games returns a ranked game list, so a combined wording can point an agent to either tool. |
Compare the field