The standard HackerOne approach: sort by pool size, target the biggest bounty. The V13 approach: which program teaches the most reusable fleet lesson? Crypto.com has a $2M pool AND KCF 9 — it wins on both axes. xAI has a $20K pool but also KCF 9 — because attacking an AI's intelligence layer IS what PEMCLAU faces. That systemic reuse is worth more than any bounty.
The V13 KCF lens also changes HOW we approach findings. Access control = S5 policy failure (not just a bug). Price manipulation = S4 intelligence compromise (same organ as GMX oracle). Novel framing beats standard PoC — especially at programs that see hundreds of standard submissions.
| Program | Pool | V13 KCF | VSM | ME-COLI | Fermentation | Why |
|---|---|---|---|---|---|---|
| Crypto.com | $2M | 9 | S3/S4 | L4/L5 | LAB | Exchange logic = S3 control failure |
| xAI | $20K | 9 | S4 | L6 | Acetic | AI system = S4 intelligence layer |
| Stripe | $25K | 8 | S3 | L4 | LAB | Payment = metabolic L4 flow |
| Robinhood | $50K | 8 | S3 | L4 | E.coli | Brokerage timing = sequencer analog |
| Airbnb | $31K | 6 | S2 | L3 | Acetic | Trust/access model |
| DoorDash | $12K | 6 | S2 | L3 | E.coli | Delivery coordination |
$2M pool = highest in H1 fintech. KCF 9 because the exchange logic maps directly to the SSAF immune taxonomy. Every finding in Crypto.com enriches the same PEMCLAU collection as the SSAF findings.
AI system = S4 INTELLIGENCE compromise. This is the most V13-aligned H1 target: attacking an AI's intelligence layer is exactly what PEMCLAU faces as a threat model. Every xAI finding = direct fleet immune memory.