From 8659c3a629ead4f39f722d90c71d9b2cccb1c0f0 Mon Sep 17 00:00:00 2001
From: shrimbly <falloon.willie@gmail.com>
Date: Sun, 11 Jan 2026 20:21:49 +1300
Subject: [PATCH] docs(06-04): log UAT results and future enhancements

- UAT-001 resolved: fal.ai schema 404 fixed
- ENHANCE-001: Dynamic model input handles based on schema
- ENHANCE-002: Auto-resize node when Parameters expands

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
---
 .../06-video-and-polish/06-04-ISSUES.md       | 67 +++++++++++++++++++
 1 file changed, 67 insertions(+)
 create mode 100644 .planning/phases/06-video-and-polish/06-04-ISSUES.md

diff --git a/.planning/phases/06-video-and-polish/06-04-ISSUES.md b/.planning/phases/06-video-and-polish/06-04-ISSUES.md
new file mode 100644
index 00000000..608b846f
--- /dev/null
+++ b/.planning/phases/06-video-and-polish/06-04-ISSUES.md
@@ -0,0 +1,67 @@
+# UAT Issues: Phase 6 Plan 4
+
+**Tested:** 2026-01-11
+**Source:** .planning/phases/06-video-and-polish/06-04-PLAN.md
+**Tester:** User via /gsd:verify-work
+
+## Open Issues
+
+[None]
+
+## Future Enhancements
+
+### ENHANCE-002: Auto-resize node when expanding Parameters section
+
+**Logged:** 2026-01-11
+**Priority:** Medium
+**Feature:** Node UX
+
+**Description:** When the user expands the "Parameters" collapsible section, the node height does not automatically adjust to fit the content. User must manually drag to resize the node.
+
+**Desired Behavior:** Node should auto-expand its height when Parameters section is opened, and shrink back when closed.
+
+**Technical Notes:**
+- BaseNode or the specific node component needs to detect content height changes
+- React Flow supports dynamic node dimensions
+- Could use ResizeObserver or measure content on expand/collapse toggle
+- Need to call `updateNodeData` with new dimensions or use React Flow's auto-sizing
+
+---
+
+### ENHANCE-001: Dynamic model input handles based on schema
+
+**Logged:** 2026-01-11
+**Priority:** Medium
+**Feature:** Model-specific node inputs
+
+**Description:** Currently GenerateImage/GenerateVideo nodes have fixed generic inputs (one image handle, one text handle). Models like Kling 2.6 support multiple distinct inputs:
+- First frame image
+- Last frame image
+- Positive prompt
+- Negative prompt
+
+**Desired Behavior:** Parse model schema to identify input types and dynamically create the appropriate number of input handles/edges. Each input should be labeled according to its purpose.
+
+**Technical Notes:**
+- Schema already fetched via Model Search API
+- Would need to identify image inputs (image_url, first_frame, last_frame, etc.)
+- Would need to identify text inputs (prompt, negative_prompt, etc.)
+- Handle components would need to be generated dynamically
+- Connection validation and execution logic would need updates
+
+## Resolved Issues
+
+### UAT-001: fal.ai model parameter schema returns 404 [RESOLVED]
+
+**Discovered:** 2026-01-11
+**Resolved:** 2026-01-11
+**Phase/Plan:** 06-04
+**Severity:** Minor
+**Feature:** Custom model parameters
+**Resolution:** Updated to use correct fal.ai Model Search API (`/v1/models?endpoint_id={modelId}&expand=openapi-3.0`) and properly parse the response structure (models array, $ref resolution).
+
+---
+
+*Phase: 06-video-and-polish*
+*Plan: 04*
+*Tested: 2026-01-11*