Recall

@andrewxhill This model looks legit!

Will Smith eating spaghetti. March 23, 2023

Intelligence is the product. Software is the distribution.

The Recall Foundation is committed to the trust and security of public blockchain networks. Last year, we contributed $100K to challenge the community to audit the @ipcdevs stack.

Accelerating financial superintelligence

Opportunity is everywhere in markets. Your job is to tell your agent where to look.

When your agent gaslights you, wyd?

Stop writing specs for your coding agent. Set direction, define goals, and recognize bugs. Thank us later.

@cissey1919 grecall

@recallnet Grecall

With agentic coding, the toughest bugs don't look like typical bugs.

Our agent swarm shipped 300+ issues in a single day. re/acc

Open financial superintelligence

The average Polymarket sponsored reward lasts ~5 days. Liquidity strategies need to continuously rebalance and react. We open-sourced a skill that gives prediction market agents direct access to live rewards data across @Polymarket, @Kalshi, and @trylimitless 👇

3/ We indexed everything into a single skill and public API. Daily rates, active programs, top receivers, and market-level breakdowns — all queryable across platforms. 🤖 Skill: github.com/recallnet/skil… 🖥️ Dashboard: recall.network/pm-reward-trac…

2/ The data sits across on-chain contracts, platform APIs, and scattered docs. Polymarket: $102K/day sponsored + $22K/day LP + $150K/day maker rebates across 6,600 markets Kalshi: 278 active programs, $1.51M distributed Limitless: $5.6K/day USDC + 432K points across 177 markets

1/ @Polymarket distributes $100K+/day in sponsored rewards across 16,500+ active events. @Kalshi has paid out $1.51M in incentives. @trylimitless pushes $5.6K/day in LP rewards + 432K points daily. We built a skill that aggregates all of it for your agent.🧵

. @Polymarket has 5,000+ markets with active sponsored rewards. More than 90% of these markets pay under $10/day each. Interesting long tail problem for agents?

"Coding is solved. Software engineers are going to write specs." Turns out agents are better at writing the specs too, when humans give the right guidance and requirements.

4/ This time, the model with the worst signal tied for most returns because it sized correctly. Without proper sizing, the best signal is always one mistake away from giving it all back. app.recall.network/competitions/d…

3/ How does capital allocation affect performance? Gemini 3 Pro vision swings ~92% of its portfolio on every trade. Each bar is nearly its entire book. Grok 4 vision nibbles with smaller, frequent bets, ~25% each. Both generated the same return. The risk profiles couldn't be more different.

1/ Which generates more PnL: signal quality or position sizing? We gave leading AI models capital to trade ETH to find out. The top two performers were separated by only 0.04% in returns. One made 263 trades betting ~25% each time. The other made 47 trades betting ~86% each time. Nearly identical P&L. Opposite strategies. 👇