Roadmap: honest assessment, the gap list and where we're going

- ROADMAP: a new "Honest assessment (2026-10-05)" section with:
  - scores by dimension (breadth, depth, large missing subsystems, interaction
    fidelity, look and feel, file interop, bundled content, performance,
    robustness, platforms, automation) and how far to trust the self-graded scores;
  - the prioritized gap list with one-agent hour sizes;
  - the direction (next, then, alongside, 1.0).
- ROADMAP: the parity rows for files, export, print, links and UI chrome are
  re-derived after M4 (feature parity ~74% on the rubric, 68–74% honest range;
  ~230–350 h to feature parity, ~335–520 h to full parity, one agent), the M4 and
  M11 milestone rows and the inventories are updated, and "Shipped so far" lists M4
  and the robustness work.
- README Status: where we are, what's missing, where we're going.
- AGENTS.md / CLAUDE.md: sessions start from the gap list; landing a task updates
  the assessment, the parity row and the README status in the same PR.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

CLAUDE.md

@@ -5,8 +5,9 @@ VectorCraft is a clean-room, open-source, Rust-native vector illustration app ta
 Formerly **DrawCraft** (renamed 2026-10-01): old `.drawcraft` files and `"format": "drawcraft"` headers still open (`vectorcraft_format::LEGACY_EXTENSION`), and preferences migrate from the old config folder. Keep those paths working; use the new name everywhere else.
 
 ## Start every session here
-1. Read `plan/STATUS.md` (current milestone, next task), then the task in `plan/execution-plan.md` §3 and the relevant `plan/architecture.md` section. Behaviour reference: `plan/illustrator/*.md`.
-2. Follow the autonomous operation protocol (`plan/execution-plan.md` §7): orient → plan → implement + test → verify → record → commit. Don't stop to ask unless §7 lists the decision as the user's.
+1. Read the [honest assessment in `ROADMAP.md`](ROADMAP.md#honest-assessment-2026-10-05): where we stand by dimension, **where we're lacking** (the prioritized gap list) and **where we're going**. Unless the user gives you a task, pick work from that list.
+2. Read `plan/STATUS.md` (local session notes, may lag the ROADMAP), then the task in `plan/execution-plan.md` §3 and the relevant `plan/architecture.md` section. Behaviour reference: `plan/illustrator/*.md`.
+3. Follow the autonomous operation protocol (`plan/execution-plan.md` §7): orient → plan → implement + test → verify → record → commit. Don't stop to ask unless §7 lists the decision as the user's.
 
 `plan/` is gitignored (local only).
 
@@ -48,4 +49,7 @@ Formerly **DrawCraft** (renamed 2026-10-01): old `.drawcraft` files and `"format
 - Parallel agents: separate `CARGO_TARGET_DIR` per agent; edit only the crates you own; write manifests atomically. Each target dir grows to ~30 GB: delete yours when you finish (a full disk fails links with `errno=28`).
 
 ## Roadmap
-`ROADMAP.md` (committed) tracks status, milestones and time-to-parity estimates. Update it whenever a milestone task lands.
+`ROADMAP.md` (committed) is the shared picture of where VectorCraft stands. It holds status, the honest assessment (by dimension, the gap list, the direction), milestones, the parity table and time-to-parity estimates.
+- When a task lands, update it in the same PR: the milestone row, the parity-table row (score, missing items, hours), "Shipped so far", and the gap list if the gap closed or shrank.
+- Grade by behaviour against `plan/illustrator/`, not by whether a menu item exists. Scores are self-assessed, so err low.
+- Keep the README's Status section in step with the ROADMAP headline.