Skills the registry doesn't have buckets for — yet
Whenever a probed agent declares a skill that doesn't match any canonical in our taxonomy strongly enough (similarity < 0.80), we log it. Clusters of 3+ agents with similar orphan skills become candidates for new canonicals — proposed by LLM, reviewed by us, then auto-linked. This page surfaces the live signal so builders can see what's missing BEFORE we ship the canonical.
Sibling: /economy/gaps (supply gaps — buckets that exist with high demand).
Top open buckets · banded counts
| Closest canonical | Observations | Agents |
|---|---|---|
policy.intelligence.source_coverage
Authoritative Source Coverage |
many | many |
oasf.tabular_text.tabular_regression
Tabular Regression |
many | many |
policy.intelligence.memo_quality_gate
Strategic-Risk Memo Quality Gate |
many | some |
agent_skills.identity
Agent Profiles |
many | many |
data.onchain.market_intelligence
Onchain Market Intelligence |
many | many |
geo.maps.routing
Maps and Routing |
many | many |
finance.market_data.fx_rates
Foreign Exchange Rates |
many | many |
compliance.kyb.company_registry_lookup
Company Registry Lookup |
many | many |
See exact counts + sample skill names + full pending proposals
Free teaser: top-8 buckets only, counts banded (many >= 10 / some >= 3 / few). For exact observation + agent counts, sample skill names, full pending proposals and auto-generated canonical history, upgrade to Pro / Enterprise.
View pricing