docs(research): tx-ordering strategy + builder matrix + Phase 1 staged issue bodies

[0xfandom]

Summary

• Tx-ordering strategy research: full per-class economics, latency, risk, policy, and brand analysis for backrun / frontrun / sandwich (out-of-scope notes for JIT-LP and CEX-DEX).
• Builder matrix: per-builder endpoint, auth, bundle-shape, ordering policy, OFAC posture, and geographic POPs for Flashbots, Titan, Beaverbuild, rsync, Eden, BuilderNet, Manifold, bloXroute, plus Agnostic / Ultrasound relays. Fan-out recommendations for backrun vs frontrun vs OFAC-touching bundles.
• Decision: backrun-only, MEV-Share v2 as the priority private-flow surface, parallel fan-out to Beaver / Titan / rsync / BuilderNet / Flashbots for public-mempool victims. Sandwich is policy-gated and economically negative for the median operator in 2025-2026; frontrun is dominated by sandwich on every axis except policy.
• Staged issue bodies for Phase 1 Part A (feat(mempool-backrun): Phase 1 Part A — Rust validate_backrun revm sim + gRPC publish #138 Rust validate_backrun + gRPC), Part B (feat(mempool-backrun): Phase 1 Part B — Go executor mempool bundle builder + fan-out #139 Go executor mempool bundle builder), Part C (feat(mempool-backrun): Phase 1 Part C — risk gates + shadow rollout + live validation #140 risk gates + shadow rollout) live in docs/issues/staged/. Issues themselves are already filed on GitHub and cross-linked.

Files Changed

File	Purpose
docs/research/tx-ordering-strategy.md	Top-level skeleton + research questions + deliverables index
docs/research/strategy-class-analysis.md	Per-class deep dive (backrun / frontrun / sandwich), comparison table, recommendation
docs/research/builder-matrix.md	Per-builder canonical comparison table + deep dives + submitter.go fan-out implementation notes
docs/issues/staged/phase1-part-a-rust-validate-backrun.md	Full body for #138
docs/issues/staged/phase1-part-b-go-mempool-bundle-builder.md	Full body for #139
docs/issues/staged/phase1-part-c-risk-gates-shadow-rollout.md	Full body for #140

Acceptance Criteria

• Strategy decision recorded with quantified per-class analysis
• Builder matrix populated for the 8 active builders + 2 relays
• Fan-out routing matrix per bundle class (backrun / frontrun / OFAC-touching) documented
• Phase 1 Part A / B / C issue bodies drafted in-repo
• Phase 1 Part A / B / C GitHub issues filed and cross-linked

Test plan

• git status clean — only the 6 new docs files committed
• No code paths modified — no build / test re-run required
• Issue cross-links resolve on GitHub: feat(mempool-backrun): Phase 1 Part A — Rust validate_backrun revm sim + gRPC publish #138 → feat(mempool-backrun): Phase 1 Part B — Go executor mempool bundle builder + fan-out #139, feat(mempool-backrun): Phase 1 Part C — risk gates + shadow rollout + live validation #140; feat(mempool-backrun): Phase 1 Part B — Go executor mempool bundle builder + fan-out #139 → feat(mempool-backrun): Phase 1 Part C — risk gates + shadow rollout + live validation #140

Linked issues

• feat(mempool-backrun): Phase 1 Part A — Rust validate_backrun revm sim + gRPC publish #138 — Phase 1 Part A — Rust validate_backrun revm sim + gRPC publish
• feat(mempool-backrun): Phase 1 Part B — Go executor mempool bundle builder + fan-out #139 — Phase 1 Part B — Go executor mempool bundle builder + fan-out
• feat(mempool-backrun): Phase 1 Part C — risk gates + shadow rollout + live validation #140 — Phase 1 Part C — risk gates + shadow rollout + live validation

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
aether	Ready	Preview, Comment	May 18, 2026 4:15pm
aether-63xv	Ready	Preview, Comment	May 18, 2026 4:15pm