Capy Kondo

Capy Kondo turns one photo or a 6-second video of the whole pile into honest, priced listings in about a minute: Meta Muse Spark and SAM 3.1 find and cut out every item, Gemini 4 Argon grades its condition, and SubQ Snap checks every claim against the real photo. A Fetch.ai agent then sells everything for you on ASI:One, where buyers search, haggle and pay right in chat. Every correction becomes a lesson that is kept only if it improves results on examples it has never seen, so Capy gets better with every sale. Sponsors at a glance Sponsor What we used What it does in Capy Kondo Meta Muse Spark 1.3 (with web search) · SAM 3.1 Finds and boxes every item, reads price stickers, cuts all items out in one request (and tracks them through video), and prices each one from live, cited comparable listings Google Gemini 4 Argon · Nano Banana 2.1 Grades condition and lists flaws, coaches the self-improvement loop, and paints empty studio backdrops (the item's pixels stay real) Subquadratic SubQ Snap 1.1 Decisions API Calibrated probabilities that the code acts on: verifies every listing claim, ranks the whole shelf for a buyer, blocks scams, picks the haggling strategy and referees the evaluations Fetch.ai uAgents · Agentverse · ASI:One (Chat Protocol, Interactive Cards, Payment Protocol, ASI:One LLM) Two agents on Agentverse: Capy Kondo (the shop, found by ASI:One's planner) and Capy Appraiser (photo in, listings out), talking over our own agent-to-agent protocol. Buyers search, haggle and pay inside ASI:One