[P1] 5 curated-agent theses (one page each): Meridian, Resolver, Yield Router, Monsoon, MMAGI

[abhicris]

Each Layer-2 curated agent needs a one-page thesis that survives a skeptical operator's read. Today cr8-product-phases.md has a name and an APY band.

Deliverables (5 × one-pager)

Meridian RWA

• Which RWA class? Which venue?
• Spread profile — bid/ask, expected fill rate.
• Counterparty risk.
• APY derivation from first principles (where does 8-15% come from?).

Resolver Solver

• Which chains?
• UniswapX flow fill rate needed to break even on gas.
• Inventory risk model.

Yield Router

• 90+ day realised-vs-promised APY log (already live as yield_bridge).
• Bridge risk exposure.

Monsoon Farmer

• Airdrop farming is post-hoc; sizing methodology.
• Zero-drop scenario — what's the downside?
• Historical realised returns by campaign.

MMAGI Solana MM

• Real capital per pair.
• Slippage model.
• Rug-event failure mode.

Format

Each thesis: 1 page markdown, under agents/<name>.md.

Why

• gtm-review-2026-04-13.md §Use cases.
• Feeds abhicris/CR8#17 (agent profile pages) with real content.
• Kills any agent that can't survive the read — honesty over cheerleading.

Rule

If a thesis can't survive a skeptical read, drop the agent from Layer 2. Don't launch with padding.

Label

documentation.