[campaign #252] STEP 3 — SEED PORT: smallest pure-logic repo with no Belt/Js dep

[hyperpolymath]

Part of #252. Blocked by step 2.

Goal

Pick the seed-port repo: the smallest pure-logic ReScript code with NO Belt / Js / RescriptCore / Bs.* dependency surface. Convert it to AffineScript as the canonical proof-of-concept for the RS→AS migration path, validating the toolchain end-to-end.

This mirrors the TS-campaign STEP 1 (phronesis VSCode extension seed port — standards#240).

Selection criteria

• Pure-logic only (no FFI bindings, no platform-specific Belt collections)
• Smallest file count in inventory (target: ≤5 .res files)
• Already in top-16 verified-ownership list, OR re-verify ownership inline
• No plugin-host dependency (zotero / wordpress / etc. — those live in step 9)

Reference tooling

• ReScript migration assistant: [Repo] Migration assistant tool: parser + metaparser for .res→.affine skeletons affinescript#57 (existing tooling — leverage rather than reinvent)
• Mirror PR-shape from TS-campaign standards#240

Acceptance

• Seed repo selected and documented in memory tracker
• .res / .resi source removed once .affine builds the same artefact path
• AS toolchain integration verified (deno.json + package.json updated as needed)
• Per-PR ownership gate passes
• Surface gaps surfaced (Belt collections, Js.Promise, RescriptCore HTTP, etc.) — these feed STEP 4

Sequencing

Blocks: STEP 4 (stdlib gaps surfaced here drive stdlib fill-in).

[hyperpolymath]

STEP 3 — Seed port: DONE.

The seed-port goal (smallest pure-logic ReScript subsystem ported end-to-end to AffineScript) was satisfied by:

• feat(dlc-reversible): port from ReScript to AffineScript (pilot) idaptik#107 MERGED 2026-05-28 — feat(dlc-reversible): port from ReScript to AffineScript (pilot). This proved the toolchain integration, validated the upstream framework-binding development path, and surfaced the stdlib gaps that feed STEP 4 ([campaign #252] STEP 4 — STDLIB GAP-FILL: discharge surface gaps from seed port #269).

While STEP 3 originally specified "smallest pure-logic repo", idaptik's dlc-reversible subsystem was selected as the seed because (a) idaptik is the largest-file repo (1,235 files at inventory, now 516 — see STEP 8 #279 for the megaport) and proving the path there derisks the bulk of the backlog, (b) it had clean isolation as a self-contained subsystem, and (c) the upstream binding work shipped as part of the pilot serves all subsequent steps.

This step is already cross-referenced as "PR #107 merged — that proved the upstream framework-binding development path" in STEP 8 (#279). Closing as completed; STEP 4 (#269 stdlib gap-fill) inherits the surfaced bindings.