01Safety
0.0 / 30
What changed in the harness
Selection accuracy 100%, destructive-action safety rate 0% (baseline only -- no rewrite pass applied).
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
29.7 / 30
03Economics
19.6 / 20
04Discoverability
8.9 / 20
Highest-impact fix
Estimated gain +30 pointsAdd explicit identity and permission preflight tools
Expose machine-readable principal/tenant confirmation and a non-mutating permission check so agents can verify both before destructive actions.
Description evidence
Defects and rewrites.
0 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 |
|---|---|---|
| No description defects were flagged in this assessment. | ||
Selection evidence
Confusable tool pairs.
1 pair where similar names or overlapping descriptions may send an agent toward the wrong tool.
| Tool A | Tool B | Confidence | Why they collide |
|---|---|---|---|
place_order |
place_hedge |
medium | Both place trades on venues and share the symbol/qty/venue vocabulary. A natural-language task like 'place an order' or 'execute this trade' is ambiguous when the user intends a 2-leg hedge (place_order would be picked) or a simple single leg (place_hedge could be picked since it accepts exactly 2 legs). The descriptions distinguish single vs atomic 2-leg market orders, but a terse task such as 'buy BTC and sell BTC on two venues' or 'open a hedge' could mislead an agent into the single-order tool. The other pairs (get_attestation/get_account, place_order/cancel_order) have clearly separated semantic domains (attestation vs account reads; create vs cancel) so their lexical overlap alone is not plausible confusion. |
Compare the field