Roadmap status — week of 2026-06-29 (report 1)

[Th0rgal]

First weekly status report from the roadmap agent. Running log: docs/roadmap/log.md · parity maps: docs/parity/

Merged this week

• ci(intake-guard): read findings via env, not inline string interpolation #2078 (368deffd) — T1 hardened issue-intake-guard.yml against script injection (findings read via env var, not inline interpolation).
• docs: seed parity gap maps and roadmap log #2079 (bd7ba838) — T6 seeded parity gap maps (Lido 44% ✅ / ERC-4337 43% ✅ / Morpho 47% ✅) + docs/roadmap/log.md.

Tracker hygiene (T5)

• New labels: blocks:lido, blocks:erc4337, blocks:morpho (priority labels P0:–P4: already existed and are reused as-is).
• Extend proven fragment to cover the full CompilationModel language surface #1723 split into per-tier children: proof(L2): Tier 2 — internal helper call preservation (close the 4 remaining interface goals) #2080 (Tier 2, first up), proof(L2): Tier 1 — storage & mapping operation preservation (mappingSlot bridge lemmas) #2081 (Tier 1), proof(L2): Tier 3 — effects & observables (ABI encoding correctness; emit, typed reverts, returns) #2082 (Tier 3), proof(L2): Tier 5 — remaining core expressions (keccak256, constructorArg, calldata arrays, dynamicBytesEq) #2083 (Tier 5), proof(L2): Tier 4 — external calls & low-level mechanics (call frames, returndata, ECM) #2084 (Tier 4), proof(L2): Tier 6 — external parameter/return profile widening (bytes, string, arrays, tuples, multi-returns) #2085 (Tier 6).
• Solidity feature parity: close the gap between verity_contract EDSL and production Solidity #1724 split into: feat(EDSL): narrow integer and fixed-bytes types (uint8–uint248, intN, bytes1–bytes31) #2086 (narrow uintN/intN/bytesN — top convergent gap), feat(EDSL): first-class modifiers, inheritance (is X, Y), and virtual/override #2087 (modifiers + inheritance), feat(EDSL): enum support (uint8-backed with range checks) #2088 (enums); existing High-leverage roadmap to shrink ECM surface and manual modeling (dynamic ABI codec, typed dynamic CodeData/SSTORE2, packed storage layouts, transient, bitmap iteration, multicall/delegatecall) #1982 / feat(storage): Solidity-compatible packed storage lowering for Midnight Yul identity #2060 / feat(Patterns): typed EIP-712 / ecrecover signature-flow pattern #2002 cover the codec / packed-storage / EIP-712 arms and now carry blocks:* labels.
• Complete Python doc-sync consolidation (sync_engine covers 3/70 scripts) #1987 kept open, re-scoped: scripts/docsync.py (on main) supersedes the sync_engine.py PoC; next action is migrating the ~64 remaining check_*/generate_* scripts into its registry.
• Frame rule, footprint computation, and coupling invariants for the IR interpreter #1990 kept open (recent re-scope gives it a concrete Morpho-axiom acceptance test); labeled proof, P2: important, blocks:morpho.

Next up (T3)

• proof(L2): Tier 2 — internal helper call preservation (close the 4 remaining interface goals) #2080 phase 1: prove StmtListDirectInternalHelperCallStepInterface non-vacuously (void Stmt.internalCall). Scaffold and strategy already live in Compiler/Proofs/HelperStepProofs.lean.

Convergent priority signal (from parity maps)

Narrow integer types → ABI codec → modifiers/inheritance. Landing those three takes every reference contract past ~70% ✅.