GPT-5 finding
Growth skill uses Basescan tokenholderlist with offset=1 to derive holder count, which does not return a total and will undercount
highbughigh
- skills/vvvkernel-growth/SKILL.md:24-26
The tokenholderlist endpoint returns a paginated list of holders, not a total count. With offset=1 it returns only one holder entry, not the total number of holders. Using this to derive holder count will produce incorrect metrics and downstream recommendations.
Recommendation
- Consult Basescan’s API for an endpoint that provides holder totals (e.g., a token stats/info endpoint) or paginate through tokenholderlist and accumulate until exhaustion (with rate limits/keys). - Alternatively, use an indexed data provider (Covalent, Alchemy Transfers API, Dune) that exposes holder counts. - Update the spec to reflect the correct method and include API key usage if required.