refactor: move pm adapter out of layout engine

[VladaHarbour]

No description provided.

[linear-code]

SD-3222

[github-actions]

Agent docs audit

Found deterministic findings on 2 changed agent-doc item(s).

AGENTS.md (101 lines)

• 2 broken path ref(s)

Broken path refs:

• core/layout-adapter/
• src/editors/v1/core/layout-adapter

packages/layout-engine/AGENTS.md (167 lines)

• 2 broken path ref(s)

Broken path refs:

• core/layout-adapter/converters/
• src/editors/v1/core/layout-adapter

---

Deterministic L1 only: no AI, no Bash, no secrets. Semantic L2/L3 audit runs weekly on main. Policy: agent-docs-policy.md.

[cubic-dev-ai]

cubic analysis

4 issues found across 213 files

Linked issue analysis

Linked issue: SD-3222: Rendering: pm-adapter

Status	Acceptance criteria	Notes
✅	pm-adapter lives outside layout-engine (packages/pm-adapter)	File paths and package metadata were updated to move the adapter out of layout-engine into packages/pm-adapter.
✅	Layout runtime packages do not import concrete adapters at runtime (Guard H)	Architecture tests were extended (Guard H) and many runtime consumers were changed to depend on the adapter contract rather than importing pm-adapter directly.
✅	Super-editor presentation path uses @superdoc/layout-adapter registry instead of direct toFlowBlocks imports	Presentation/editor code no longer imports toFlowBlocks directly and instead uses getLayoutDocumentAdapter()/register.
✅	Default ProseMirror adapter registers via @superdoc/pm-adapter/register	A register entry was added in the pm-adapter package and the register import is pulled in at startup/test setup to ensure a default adapter is present.
✅	Layout integration tests convert documents through the adapter contract helper	Tests now route conversions through a to-flow-blocks helper that calls getLayoutDocumentAdapter().toFlowBlocks, and many tests were updated to use it.

<sub>Partial review: This PR has more than 50 files, so cubic reviewed the highest-priority files first. During the trial, paid plans get a higher file limit.
You can try an ultrareview to bypass the file limit, comment @cubic-dev-ai ultrareview. Learn more.

Fix all with cubic | Re-trigger cubic</sub>

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[cubic-dev-ai]

1 issue found across 214 files (changes from recent commits).

Prompt for AI agents (unresolved issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="packages/layout-engine/tests/src/architecture-boundaries.test.ts">

<violation number="1" location="packages/layout-engine/tests/src/architecture-boundaries.test.ts:21">
P1: V1_ADAPTER_ROOT path has one extra `..` level — it resolves to `<monorepo-root>/super-editor/...` instead of `packages/super-editor/...`. Since `collectRuntimeSources` silently returns an empty array for a non-existent directory, all Guard B and Guard C tests will pass without actually scanning the v1 adapter source, giving false confidence in the architecture boundaries.</violation>
</file>

<sub>Tip: Review your code locally with the cubic CLI to iterate faster.
Partial review: This PR has more than 50 files, so cubic reviewed the highest-priority files first. During the trial, paid plans get a higher file limit.
You can try an ultrareview to bypass the file limit, comment @cubic-dev-ai ultrareview. Learn more.

Fix all with cubic | Re-trigger cubic</sub>

[cubic-dev-ai]

P1: V1_ADAPTER_ROOT path has one extra .. level — it resolves to <monorepo-root>/super-editor/... instead of packages/super-editor/.... Since collectRuntimeSources silently returns an empty array for a non-existent directory, all Guard B and Guard C tests will pass without actually scanning the v1 adapter source, giving false confidence in the architecture boundaries.

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At packages/layout-engine/tests/src/architecture-boundaries.test.ts, line 21:

<comment>V1_ADAPTER_ROOT path has one extra `..` level — it resolves to `<monorepo-root>/super-editor/...` instead of `packages/super-editor/...`. Since `collectRuntimeSources` silently returns an empty array for a non-existent directory, all Guard B and Guard C tests will pass without actually scanning the v1 adapter source, giving false confidence in the architecture boundaries.</comment>

<file context>
@@ -15,7 +15,10 @@ import fs from 'node:fs';
+// SD-3222: the v1 ProseMirror adapter now lives inside @superdoc/super-editor
+// (it is v1 SuperEditor's projection from hidden PM state into FlowBlock[]),
+// not in a standalone layout-engine package.
+const V1_ADAPTER_ROOT = path.resolve(__dirname, '../../../../super-editor/src/editors/v1/core/layout-adapter');
 
 // ---------------------------------------------------------------------------
</file context>

Suggested change

	const V1_ADAPTER_ROOT = path.resolve(__dirname, '../../../../super-editor/src/editors/v1/core/layout-adapter');
	const V1_ADAPTER_ROOT = path.resolve(__dirname, '../../../super-editor/src/editors/v1/core/layout-adapter');