4 routines · launches, KOL engagement, betting

AI agents that launch, engage, and bet for you

Four routines built on the platform's automation engine, each running on your own credits and your own platform wallet. Configure one and it runs on a schedule you set, no dashboard babysitting required.

Launch a meme coin

On demand

Give it a name, ticker, and theme. It generates the art, mints on Solana or Base from your platform wallet, and reports back the live coin page.

Run onceScheduledWebhook

Real cost per run

~111-130 credits

Agent run (10) + generated art (1-20) + mint (pumpfun_create_coin / base_create_coin, 100).

Configure this automation

Launch a meme coin from the news

News-driven

Scans breaking crypto-adjacent stories and only launches when something genuinely qualifies. Most runs launch nothing, on purpose.

HourlyScheduledWebhook

Real cost per run

~112-135 per launch credits

Agent run (10) + news scan (twitterio_query, 1 per topic) + art (1-20) + mint (100). 0 credits on a run that launches nothing.

Configure this automation

KOL engagement for a coin

Ongoing

Watches your coin's cashtag and community, likes and replies to genuine conversation within caps you set, and can post original updates.

HourlyScheduledWebhook

Real cost per run

~25-320 credits

Agent run (10) + reads (twitterio_query, 1 each) + likes (2 each, capped) + replies/posts (80 each, capped).

Configure this automation

Autonomous prediction betting

Ongoing

Finds new prediction markets worth a credit-based bet within your exposure cap, sets take-profit and stop-loss on every position, never sells itself.

HourlyScheduled

Real cost per run

~10 + stakes credits

Agent run (10). Bets stake credits directly, capped by your max stake per bet and total exposure cap (no per-tool fee).

Configure this automation

Same wallet, same credits, no new setup

Every launch and trade signs with your platform wallet, the same one you already use to launch coins by hand.

Credits burn exactly as they would if you did each step yourself: an agent run, plus one charge per tool call it actually makes.

Configured in the workspace's routine builder, the same one behind every other automation on the platform.