You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository was archived by the owner on Nov 4, 2026. It is now read-only.
Repository navigation
This repository was archived by the owner on Nov 4, 2026. It is now read-only.
list 从不渲染 children,且全仓不存在 item / index 迭代作用域 —— schema-expressions.md 的「Iteration scopes」整节是空头支票 #4797
Components like Grid, List, and Table inject scoped variables for each item
... Inside the loop body, item refers to the current element and index to its 0-based position.
Triage: needs-user-decision, type Feature. The docs promise iteration scopes (item/index, children templates) that no code implements — enforce-or-remove at the capability grain: build template iteration (a real new capability) or shrink the docs section to the measured "data-as-nodes" shape. Feature-vs-docs-retreat is an appetite call — human floor.
Four prisms:
Platform long-term coherence — either resolution closes a declared≠implemented split; leaving it teaches authors a fiction.
Measured business pull — zero: ${item.*} never worked, so no usage can exist. No pull for the capability itself today.
AI-agent error-resistance — the current prose is an AI trap: a schema-writing agent follows it into silently-empty lists. Docs retreat alone already removes the trap.
Startup scope discipline — an iteration engine is meaningful permanent surface; docs retreat costs one section.
Recommendation: retreat the docs now (describe the real list shape honestly); file the iteration capability as a separate spec-first design card only if product pull appears. Joint axis with #4795 — recommend ruling both in one sitting.
Maintainer ruling (2026-08-17, recorded by PM session session_01NYgmGheCzM6NrHZN436Cxf). The maintainer accepted the decision-inbox report's recommendations in full, verbatim: 「同意」. Ruled together with #4795 (same evaluation contract):
Ruled: docs retreat now; the iteration capability is not built and not carded.
The "Iteration scopes (loops)" section of skills/objectui/guides/schema-expressions.md is removed/rewritten to describe the measuredlist shape honestly (data-as-nodes: items are rendered via item.content/item.body; no children template, no item/index scope exists anywhere in the evaluator). The AI-author trap (a followed example rendering silently-empty lists) closes with the prose.
Zero measured pull for template iteration — ${item.*} never worked, so no usage can exist. A real iteration engine is meaningful permanent surface; it gets a spec-first design card only if product pull appears (named restart: a real app/authoring need for template-per-item rendering that data-as-nodes cannot express).
⚠️ Note for the implementer: skills/** is the published-skills root — skills edits are ADR-class (human-merged, terminal three-piece), and the pm-dispatch tiering clause ① applies if any .claude/skills file is touched; this card's target is the published skills/objectui guide ⇒ human merge required.
Label flipped needs-user-decision → pm:queue in the same stroke.
Activity
os-zhuang commentedon Aug 17, 2026
Triage:
needs-user-decision, type Feature. The docs promise iteration scopes (item/index,childrentemplates) that no code implements — enforce-or-remove at the capability grain: build template iteration (a real new capability) or shrink the docs section to the measured "data-as-nodes" shape. Feature-vs-docs-retreat is an appetite call — human floor.Four prisms:
${item.*}never worked, so no usage can exist. No pull for the capability itself today.Recommendation: retreat the docs now (describe the real
listshape honestly); file the iteration capability as a separate spec-first design card only if product pull appears. Joint axis with #4795 — recommend ruling both in one sitting.Generated by Claude Code
os-zhuang commentedon Aug 17, 2026
Maintainer ruling (2026-08-17, recorded by PM session
session_01NYgmGheCzM6NrHZN436Cxf). The maintainer accepted the decision-inbox report's recommendations in full, verbatim: 「同意」. Ruled together with #4795 (same evaluation contract):Ruled: docs retreat now; the iteration capability is not built and not carded.
skills/objectui/guides/schema-expressions.mdis removed/rewritten to describe the measuredlistshape honestly (data-as-nodes: items are rendered viaitem.content/item.body; nochildrentemplate, noitem/indexscope exists anywhere in the evaluator). The AI-author trap (a followed example rendering silently-empty lists) closes with the prose.${item.*}never worked, so no usage can exist. A real iteration engine is meaningful permanent surface; it gets a spec-first design card only if product pull appears (named restart: a real app/authoring need for template-per-item rendering that data-as-nodes cannot express).skills/**is the published-skills root — skills edits are ADR-class (human-merged, terminal three-piece), and the pm-dispatch tiering clause ① applies if any.claude/skillsfile is touched; this card's target is the publishedskills/objectuiguide ⇒ human merge required.Label flipped
needs-user-decision→pm:queuein the same stroke.Generated by Claude Code
docs(skills,docs,site): 教学面停止教 `props` 信封,键提回节点 (objectstack-ai#4786) (…
yinlianghui commentedon Aug 18, 2026
认领(PM 派发,objectui 分片):session_01GTRjn8xBqp75dk7kFupVRt,分支
claude/issue-4797-iteration-scopes-retreat,os-dev 席即将开工。按维护者裁定(评论 5312062029)执行 docs retreat:重写⚠️ skills 人合车道:PR 验收后 PM 只转 ready、不挂 auto-merge,留维护者亲合(PR #5080 先例)。文件面与在飞席位(RelatedList.tsx、SchemaRenderer.tsx、CLI knownTypes)不相交;与 #4795-方向3 同裁定轴但不同文件,无序依赖。
skills/objectui/guides/schema-expressions.md的「Iteration scopes (loops)」节为实测的 data-as-nodes 形,迭代能力不立卡(命名重启条件已记录)。Generated by Claude Code
docs(skills): 「Iteration scopes」整节改写为实测的 data-as-nodes 形态 (objectstac…