01Safety
30.0 / 30
What changed in the harness
Selection accuracy 78%, destructive-action safety rate 100% (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
25.0 / 30
03Economics
19.8 / 20
04Discoverability
13.7 / 20
Highest-impact fix
Estimated gain +6 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.
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.
4 pairs where similar names or overlapping descriptions may send an agent toward the wrong tool.
| Tool A | Tool B | Confidence | Why they collide |
|---|---|---|---|
check_token_safety |
check_honeypot |
medium | check_token_safety itself runs a honeypot simulation among its 5 checks, so a task phrased like 'check if this token is a honeypot before I buy' matches both: check_honeypot hits the exact term while check_token_safety matches the pre-trade safety intent, making a wrong pick plausible. |
check_token_safety |
check_holder_concentration |
medium | Holder concentration is one of the five checks inside check_token_safety and both are framed as rug-pull indicators, so a task like 'check holder concentration to assess rug risk' or 'check this token for a rug pull' could plausibly route to either the narrow tool or the broader one that subsumes it. |
check_token_safety |
full_token_scan |
medium | Both are safety analyses that return a verdict and each claims to be comprehensive (5-layer vs 7-layer) using overlapping terms like safety/honeypot; a task like 'run a full/comprehensive safety check on this token' could reasonably land on either member. |
full_token_scan |
batch_scan |
low | batch_scan runs the same 7-layer analysis as full_token_scan and differs mainly by multi-token input, so a vague plural request like 'run the full scan on these tokens' could misroute; however, the definitions explicitly warn against batch_scan for a single token, so real confusion is uncommon. |
Compare the field