Files
editor/packages/mcp/test-reports/phase8/p2-variants.md
T
Adrian PerezandClaude Opus 4.7 0b84e7b7b1 fix(mcp,editor): close URL-validation bypasses surfaced by Phase 8 P4
Phase 8 parallel validation flagged two boundaries where malicious URLs
(javascript:, file:, external http:, data:text/html, ...) could be
persisted despite the AssetUrl allowlist added in Phase 7 A7:

1. `save_scene({ includeCurrentScene: false, graph })` — the graph arg
   was treated as opaque (`z.record(z.string(), z.unknown())`) and
   written to the store without re-running AnyNode.safeParse.

2. `POST /api/scenes { graph }` in the editor API — same issue; the
   Zod `graphSchema` accepted anything object-shaped.

Fixes:
- `save-scene.ts`: when `includeCurrentScene === false`, iterate every
  node and run `AnyNode.safeParse`; collect issues and throw
  `McpError(InvalidParams, 'graph_invalid', { errors })` on any
  failure.
- `app/api/scenes/route.ts`: replace `graphSchema` with a structured
  `z.object({ nodes, rootNodeIds, collections? })` + `superRefine`
  that runs `AnyNode.safeParse` on every node. Invalid → 400 with
  detailed issue paths.

Tests:
- Added `save_scene` regression test for the P4 attack
  (item.asset.src = 'javascript:alert(1)') — expected error.
- Fixed the existing `includeCurrentScene=false` test to use a
  schema-compliant site node id (the prior `id: 'root'` now fails
  the AnyNode parse, which is the desired strict behaviour).
- Full suite: 294 pass / 0 fail.

Also adds Phase 8 test-reports/phase8/** (10 agents, ~15 scripts +
markdown reports) documenting the validation run, plus minor biome
cleanups to the Phase 5/7 test artefacts (removed stale
`// biome-ignore` suppression comments that now resolve to the
already-off `noConsole` rule).

Phase 8 result summary (10 parallel agents, stdio MCP transport with
isolated data dirs):
- P1 templates: 18/18 PASS
- P2 variants: 6/7 mutations + determinism + save + combined + error
- P3 locking: 12/12 PASS (MCP + editor HTTP If-Match)
- P4 URL hardening: fixed 2 bypasses (see above)
- P5 photo-to-scene: 6/6 PASS
- P6 Casa del Sol via save_scene: 13/13 PASS
- P7 editor HTTP API: 18/18 PASS
- P8 concurrency: 4/5 PASS, flagged 2 real filesystem-store races
  (expectedVersion CAS gap + .index.json drift under parallel writes)
- P9 edge cases: 13/13 PASS (size cap, slug safety, bad inputs)
- P10 full sweep: 37/37 PASS (30 tools + 4 resources + 3 prompts)

Known follow-ups:
- FilesystemSceneStore needs a proper lockfile / atomic CAS to fix
  the P8 concurrency bugs (low priority: single-writer MCP is the
  typical case).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-19 20:26:26 +02:00

3.0 KiB

P2 Phase 8 — generate_variants report

Generated: 2026-04-19T18:19:47.106Z Data dir: /tmp/pascal-phase8-p2 Transport: stdio (bun packages/mcp/dist/bin/pascal-mcp.js --stdio) Total run time: 32 ms

Setup

  • template: two-bedroom
  • base nodeCount: 25
  • base saved: true (id=4a051220b1e6)

Per-mutation results

# Mutation Status nodeCounts Summary
1 wall-thickness PASS [25, 25] 2 variants, nodeCounts=[25,25], min=25
2 wall-height PASS [25, 25] 2 variants, nodeCounts=[25,25], min=25
3 zone-labels PASS [25, 25] 2 variants, nodeCounts=[25,25], min=25
4 room-proportions PASS [25, 25] 2 variants, nodeCounts=[25,25], min=25
5 open-plan FAIL [23, 24] variant nodeCount 23 < min 24 (base=25)
6 door-positions PASS [25, 25] 2 variants, nodeCounts=[25,25], min=25
7 fence-style PASS [25, 25] 2 variants, nodeCounts=[25,25], min=25

Variant descriptions

  • wall-thickness: "wall thickness 0.2m", "wall thickness 0.25m"
  • wall-height: "wall height 2.7m", "wall height 3m"
  • zone-labels: "zones [Living / Kitchen, Bedroom 2, Bedroom 1, Bath]", "zones [Bath, Bedroom 1, Living / Kitchen, Bedroom 2]"
  • room-proportions: "room proportions nudged", "room proportions nudged"
  • open-plan: "open-plan", "open-plan"
  • door-positions: "doors repositioned", "doors repositioned"
  • fence-style: "no-op", "no-op"

Determinism

  • Status: PASS
  • Detail: 3 variant graphs identical (after ID normalization) across calls (wall-thickness, seed=1337)

Save path

  • Status: PASS
  • Detail: variants saved=3, list_scenes returned 4 (expected 4)
  • Variants saved in step: 3
  • list_scenes after save: 4

Combined mutation validation

  • Status: PASS
  • Detail: combined variant sceneId=18e32febadc7, valid=true, errors=0, description="wall thickness 0.1m, wall height 2.7m, zones [Bedroom 2, Living / Kitchen, Bath, Bedroom 1], room proportions nudged, open-plan, doors repositioned"
  • valid: true, errorCount: 0

Error path

  • Status: PASS
  • Detail: isError with text: MCP error -32602: scene_not_found

Totals

  • Total variants saved across the run: 4

Overall summary

Summary (≤150 words):

Per-mutation: 6/7 PASS. Determinism: PASS (identical after id normalization; forkSceneGraph regenerates ids so raw JSON can't match). Save path: PASS — 3 variants saved, list_scenes returned 4 (expected 4). Combined mutation: PASS (variant validates). Error path: PASS. Total variants saved: 4. Total variants exercised across the run: ~21.

Note: the only failing mutation is open-plan — nodeCounts [23, 24] with base=25. The spec rule >= base - 1 assumes open-plan drops only the wall node, but applyOpenPlan also drops any openings (doors/windows) attached to the removed wall — so a variant may drop 2+ nodes. The mutation itself is working correctly; the spec's lower-bound rule is tighter than the implementation.