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
28.0 / 30
03Economics
20.0 / 20
04Discoverability
15.7 / 20
Highest-impact fix
Estimated gain +4 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.
4 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 |
|---|---|---|
reverse_dns |
no_return_description |
Perform a reverse DNS (PTR) lookup on an IP address and return the hostname(s) associated with it. |
resolve_all |
name_restates_behavior no_return_description |
Resolve all common DNS record types (A, AAAA, MX, TXT, NS, CNAME) for a domain in one call and return the records found for each type. |
check_nameservers |
no_return_description |
Get the nameserver (NS) records for a domain and return the list of authoritative nameservers assigned to it. |
whois |
no_return_description |
Query WHOIS data for a domain, using IANA and following referrals to registrar WHOIS servers, and return the raw WHOIS record for the domain. |
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 |
|---|---|---|---|
dns_lookup |
reverse_dns |
medium | A task phrased as 'do a reverse DNS / PTR lookup for 8.8.8.8' is ambiguous because dns_lookup's type enum explicitly includes PTR, so an agent could call dns_lookup with type=PTR on an IP address instead of reverse_dns, even though descriptions nominally differ on domain vs IP input. |
Compare the field