[CameronBrooks11]

Several status lines had fallen behind.

• README.md kinds table: blockspec's status "design pass" becomes "specification draft 0, conformance cases and reference implementation".
• README.md intro: "alongside rowspec (the row kind)" also names blockspec (the block kind).
• DESIGN-BRIEF.md §4: the note that node identity is "barely more than a slogan at this stage" gains what 0001 decides (a node is its declared id) and what it leaves open (whether D1 to D5 refuse every identity error; the delete-against-move residual; canonical statement order).
• DESIGN-BRIEF.md §0, README.md and AGENTS.md "Before writing anything": the adversary question is recorded as settled, citing 0001. In AGENTS.md only the status wording changes; no rule changes.

Evidence for each claim

• blockspec has specification draft 0: head -2 SPEC.md in kindspec/blockspec at 3629226 prints # blockspec — specification, draft 0.
• blockspec has conformance cases and a reference implementation: conformance/cases/ and reference/blockspec/ (document.py, anchor.py) exist at 3629226, and just conform there prints 0 failure(s) across 547 case(s) in the fixture tree.
• blockspec is the block kind: blockspec's README.md says "blockspec is the block kind".
• A node is its declared id: 0001 D1, "Every id is declared exactly once, and everything else only names it", and Not prevented, "The id is the node's identity, by this profile's own rule" (grep -n '^\*\*D1\|identity, by this profile' docs/adjudications/0001-mermaid-profile.md).
• Still open whether D1 to D5 refuse every identity error: 0001 Open, "Whether some identity error passes D1 to D5 with one ordinary edit per leg beyond those named above" (grep -n 'passes D1 to D5' docs/adjudications/0001-mermaid-profile.md).
• 0001 names merges they let through, among them a delete against a move: 0001 Not prevented, "A delete against a move of the same declaration … the deleted thing comes back" (grep -n 'delete against a move' docs/adjudications/0001-mermaid-profile.md).
• Canonical statement order is not decided: SPEC.md §12 Not decided lists "A canonical statement order, which would close the delete-against-move residual" (grep -n 'canonical statement order' -i SPEC.md).
• The adversary is settled: 0001 What is decided already, "Its adversary is blockspec's (owner, 2026-10-19): this text, then conformance cases written blind from the specification it becomes, then a reference checker run against them unchanged, under a mutation gate" (grep -n 'adversary is blockspec' docs/adjudications/0001-mermaid-profile.md). 0001's status line reads "approved by the owner, 2026-10-19".

Checked and left as is

• rowspec's row "draft 0, published" is accurate: PyPI's latest rowspec is 0.4.0 (curl -s https://pypi.org/pypi/rowspec/json), and gh release list -R kindspec/rowspec shows v0.4.0, "rowspec 0.4.0 — draft 0", as latest.
• nodespec's own row and status paragraph match SPEC.md (draft 0), reference/nodespec/ and conformance/cases/ (just conform: 322 passed, 0 failed).
• §4 "Name availability" is still accurate: PyPI and crates.io return 404 for nodespec, and the npm registry returns 200.
• README.md and DESIGN-BRIEF.md carry no issue references.

Gate

just check, just test (430 passed) and just conform (322 passed) each exit 0 on this branch.