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
29.4 / 30
03Economics
19.1 / 20
04Discoverability
13.3 / 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.
3 pairs where similar names or overlapping descriptions may send an agent toward the wrong tool.
| Tool A | Tool B | Confidence | Why they collide |
|---|---|---|---|
scan_gtm_hiring_signals |
aggregate_gtm_signals |
medium | Both detect a company's GTM activity on a domain, take a bare domain, and consume Apify credits, but one returns raw job postings and the other a composite score that also runs hiring detection. A task like 'assess this company's GTM hiring signals' or 'get GTM signals for stripe.com' doesn't disambiguate whether the agent wants the job listings or the scored summary. |
resolve_linkedin_url |
resolve_company_identity |
medium | Both take a company name/domain and return the LinkedIn company URL with confidence scores in a flat Clay-ready row. A task like 'resolve this company to its LinkedIn page with a confidence score' from only the tool definitions could land on either, since resolve_company_identity also returns the LinkedIn URL and confidence; the distinction (LinkedIn-only vs canonical identity cross-check) only surfaces in edge cases. |
resolve_company_identity |
enrich_company_firmographics |
low | Both consume a bare company domain and return flat company metadata (name, domain, logos, confidence) as Clay-ready rows. A vague task like 'look up details about stripe.com' or 'enrich this company' doesn't clearly signal identity resolution versus firmographics enrichment, though name and purpose divergence (identity/cross-check vs employee/industry data) usually pulls an agent to the right one. |
Compare the field