diff --git a/.gitattributes b/.gitattributes
index 59f57df35fe962f3d1d4fba28ce2f392f27ec3ed..ff98a0ea75bae70aa3cf47e1536de693f10bfb43 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -678,3 +678,22 @@ FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-00467/task-00467.x
 FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-00468/task-00468.xlsx filter=lfs diff=lfs merge=lfs -text
 FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02018/task-02018.xlsx filter=lfs diff=lfs merge=lfs -text
 FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02733/task-02733.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03344/task-03344.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03463/task-03463.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03534/task-03534.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03666/task-03666.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05496/task-05496.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03389/task-03389.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03537/task-03537.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06116/task-06116.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02496/task-02496.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02831/task-02831.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02832/task-02832.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03286/task-03286.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03287/task-03287.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03288/task-03288.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02147/task-02147.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05962/task-05962.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06510/task-06510.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06514/task-06514.xlsx filter=lfs diff=lfs merge=lfs -text
+FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02016/task-02016.xlsx filter=lfs diff=lfs merge=lfs -text
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02016/task-02016.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02016/task-02016.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..2706eb4207aea7ca74b81c07c8aca61f8d8ea0a4
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02016/task-02016.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f873e6623232f243a4fbfd0f42fc928cda7317d3c2c9c5aee0efab2becbaaead
+size 222695
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02016/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02016/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..de4dd4a89e034375e46038a50f9446248b4c84ea
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02016/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-02016",
+  "old_fcb_id": "FCB-06160",
+  "task_id": "cross_sheet__dag_c393e9942f0f",
+  "release_layer": "Core",
+  "prompt_en": "The DefaultProbs sheet is the monthly default-probability area of this workbook, with rating and date labels like Effective Date, E-rating, S&P, Year, AAA, and notes such as DO NOT CHANGE and Default probabilities below are linked. The Counterp section is the part that still needs to be completed.\n\nPlease fill in the blank cells on Sheet \"DefaultProbs\" in column AE, under \"TABLE OF CUMULATIVE PROB OF NO DEFAULT (BY MONTH) - COUNTERP\", covering AE8:AE366. If any other visible blank rating or year fields in the workbook need to be in place before this Counterp monthly no-default column makes sense, complete those as needed while leaving the areas marked DO NOT CHANGE or DO NOT MODIFY as they are.\n\nSave the completed file as an .xlsx.",
+  "prompt_cn": "“DefaultProbs”工作表是这个工作簿里按月展示违约概率的区域，里面有 Effective Date、E-rating、S&P、Year、AAA 等评级和日期标签，也能看到 DO NOT CHANGE 以及 Default probabilities below are linked 这类说明。现在还需要补完整的是 Counterp 这一部分。\n\n请填写“DefaultProbs”工作表中 AE 列的空白单元格，位置在 “TABLE OF CUMULATIVE PROB OF NO DEFAULT (BY MONTH) - COUNTERP” 下面，范围是 AE8:AE366。如果工作簿里其他可见的空白评级或年份字段需要先补上，才能让这个 Counterp 的按月无违约概率列完整合理，也请一并补好；标有 DO NOT CHANGE 或 DO NOT MODIFY 的区域请保持不变。\n\n请将完成后的文件保存为 .xlsx 格式。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_cross",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02147/task-02147.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02147/task-02147.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..ac8bf101cab7a9b12b2871d0e8110c09798e07a0
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02147/task-02147.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da8da87a999299a2cb092f689fcb3d115e6143da04e5f450f5ce86c353509ff1
+size 199366
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02147/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02147/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..fa5b6f9efd8f8a931aeb85ae7792613f7feef40b
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02147/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-02147",
+  "old_fcb_id": "FCB-06311",
+  "task_id": "cross_sheet__dag_35bcd0a2ea5f",
+  "release_layer": "Core",
+  "prompt_en": "The workbook has a “Survival Rates” sheet laid out by rating and year, with labels like “Eff. Date,” “S&P,” “Year,” “AAA,” and notes such as “For each rating, probs by year for 15 years” and “Default probabilities below are linked.” Please complete the blank entries in the “Eff. Date” column on the “Survival Rates” sheet, specifically E5:E184. Some of the blank areas in the workbook naturally feed into other sections, so make sure the relevant date entries are in place where the sheet expects them before treating the linked default-probability area as complete. When you’re done, save the completed file as an .xlsx.",
+  "prompt_cn": "这个工作簿里有一张“Survival Rates”工作表，按评级和年份整理，能看到“Eff. Date”、“S&P”、“Year”、“AAA”等标签，以及“For each rating, probs by year for 15 years”和“Default probabilities below are linked”等说明。请补全“Survival Rates”工作表中“Eff. Date”列的空白内容，范围是 E5:E184。工作簿里有些空白区域会自然影响到其他部分，所以在确认下方链接的违约概率区域完整之前，请先把工作表需要的相关日期填好。完成后，请将文件保存为 .xlsx 格式。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_cross",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02159/task-02159.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02159/task-02159.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..344f3aa329d7b6324d921383966b53a9150f3471
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02159/task-02159.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02159/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02159/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..69878278f239f0579e8268eb7a9f88622fd93467
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02159/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-02159",
+  "old_fcb_id": "FCB-00365",
+  "task_id": "single_sheet__dag_5dd929eda419",
+  "release_layer": "Core",
+  "prompt_en": "The Example sheet is a finance workbook for a loan repayment schedule, with visible setup fields such as Annual Interest Rate, Loan Period (in months), Original Repayment Amount, and Loan Start Date. For the fiscal year close-out review, please complete the blank entries in the % Capital Outstanding column on the Example sheet, specifically cells I13:I372. Some of the monthly schedule fields in the workbook naturally need to be filled before this percentage column will make sense, so make sure the related blank areas are completed as needed. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Example 工作表是一份用于贷款还款计划的财务工作簿，表中能看到 Annual Interest Rate、Loan Period (in months)、Original Repayment Amount 和 Loan Start Date 等设置字段。为了进行财政年度结账复核，请补全 Example 工作表中 % Capital Outstanding 列的空白内容，具体位置是 I13:I372。工作簿里有些月度计划相关字段需要先补齐，这个百分比列才会合理，所以请根据需要把相关空白区域也补完整。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02190/task-02190.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02190/task-02190.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..9b251c5454d6dcd2ea173a3559881078850cdca3
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02190/task-02190.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02190/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02190/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..c2d996aa97de6a94deb45016dc807f3b10167540
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02190/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-02190",
+  "old_fcb_id": "FCB-07108",
+  "task_id": "cross_sheet__dag_c6c11e448355",
+  "release_layer": "Core",
+  "prompt_en": "The Stock sheet is set up as an item list with columns like Item #, Description, Qty, CODE, Units, Sold, Remain, Unit Price, %age, and Sale Price, and the remaining-stock entries are the part still missing. Please fill in the Remain column on the Stock sheet for rows 3 through 402, using the visible item and sales information already present in the workbook. Some of the related item fields in the workbook naturally need to be in place before the Remain column makes sense, so keep the sheet context in mind as you complete those entries. Please save the completed file as an .xlsx workbook.",
+  "prompt_cn": "Stock 工作表是一个物品清单，里面有 Item #、Description、Qty、CODE、Units、Sold、Remain、Unit Price、%age 和 Sale Price 等列，目前缺少的是剩余库存相关的 Remain 列内容。请根据工作簿中已经可见的物品和销售信息，填写 Stock 工作表第 3 行到第 402 行的 Remain 列。工作簿里一些相关的物品字段需要先保持一致，Remain 列的内容才有意义，所以请结合整个表格的上下文完成这些条目。请将完成后的文件保存为 .xlsx 工作簿。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_cross",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02192/task-02192.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02192/task-02192.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..98a9c4b2fbb30604f5bd9f61c725129a430bab03
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02192/task-02192.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02192/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02192/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..2f33d247de0ebf52e8300f95f74fff83376a0930
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02192/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-02192",
+  "old_fcb_id": "FCB-07100",
+  "task_id": "cross_sheet__dag_a2c5b9432f32",
+  "release_layer": "Core",
+  "prompt_en": "The Bid Summary Sheet workbook is organized around March Start Bids and related bid sheets, with visible sections for Dominion, Oxy, April Bid, Original Bid - February 12, 2002, and monthly bid information from March through October. On the \"Dominion Bid\" sheet, please fill in the blank entries in the \"Apr - Oct WACOG\" column, specifically L8:L38. Some of the monthly bid figures elsewhere in the workbook need to be in place before the Apr - Oct WACOG entries make sense, so use the visible workbook structure and labels to complete the Dominion Bid section consistently. Please save the completed file as an .xlsx.",
+  "prompt_cn": "这个 Bid Summary Sheet 工作簿围绕 March Start Bids 和相关投标表展开，里面能看到 Dominion、Oxy、April Bid、Original Bid - February 12, 2002，以及从 March 到 October 的月度投标信息。请在 \"Dominion Bid\" 工作表中补全 \"Apr - Oct WACOG\" 列的空白项，位置是 L8:L38。工作簿中其他月度投标数据需要先保持完整，\"Apr - Oct WACOG\" 这一列的内容才会合理，所以请根据可见的表格结构和标签，把 Dominion Bid 部分补齐。请将完成后的文件保存为 .xlsx 格式。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_cross",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02496/task-02496.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02496/task-02496.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..0fc4ed4bd2ad89c6fb0088413a61bd4d6166993b
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02496/task-02496.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb98e259a476130a1724ad1a6d01337fd73028d060b2f6b1cdd17fd072f2a20c
+size 115817
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02496/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02496/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..fd7e21584f72ba0a13fe1d182d51d87f424ff0e8
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02496/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-02496",
+  "old_fcb_id": "FCB-00929",
+  "task_id": "single_sheet__dag_2fcd35048f27",
+  "release_layer": "Core",
+  "prompt_en": "On the 12H sheet, the well report is organized by daily rows with fields like Date, Time, TP, CP, Surf CP, Sep PSI, Temp, and the Gas Sales column near the right side. Please fill in the blank Gas Sales entries in column AM, rows 3 through 242, for the 12H report. Some of the other row details on this sheet may need to be in place before the Gas Sales entries are meaningful, so keep the related daily records consistent as you complete the column. When finished, save the completed workbook as an .xlsx file.",
+  "prompt_cn": "在“12H”工作表中，这份油井报表按每日记录排列，包含 Date、Time、TP、CP、Surf CP、Sep PSI、Temp 等字段，右侧还有 Gas Sales 列。请补全“12H”报表中 AM 列第 3 行到第 242 行的 Gas Sales 空白项。该表中的一些同行记录需要先保持完整，Gas Sales 的内容才有意义，因此在补全这一列时请确保相关的每日记录一致。完成后，请将工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02706/task-02706.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02706/task-02706.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..5a421254570a1cec2bc9363f556c7c98f58ebed0
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02706/task-02706.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02706/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02706/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..c6ff34e958a0cc80f79e13cc7e892854401b31a4
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02706/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-02706",
+  "old_fcb_id": "FCB-09928",
+  "task_id": "cross_sheet__dag_36df1905653f",
+  "release_layer": "Core",
+  "prompt_en": "The \"Player Accum Earnings\" sheet is tracking each Player across the Week 1 through Week 11 earnings columns, with related ranking sections such as \"Rankings Through Week 2\" and \"Rankings Through Week 3\" visible elsewhere in the workbook. Please fill in the missing entries in the \"Total\" column on the \"Player Accum Earnings\" sheet, specifically U6:U487, so each player’s row is complete. Some of the weekly and ranking-related blanks in the workbook feed into the final player summary, so make sure the supporting entries are in place before completing the \"Total\" column. Save the completed workbook as an .xlsx file.",
+  "prompt_cn": "“Player Accum Earnings”工作表按 Player 记录了 Week 1 到 Week 11 的 earnings 列，工作簿中还可以看到 “Rankings Through Week 2”、“Rankings Through Week 3”等相关排名部分。请补全 “Player Accum Earnings” 工作表中 “Total” 列的缺失内容，范围是 U6:U487，让每位球员对应的行都完整。工作簿里有些周度和排名相关的空白内容会影响最终的球员汇总，因此在填写 “Total” 列之前，请先确保相关的支持内容已经补齐。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_cross",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02831/task-02831.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02831/task-02831.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..1e3bb3062d6747836b3b4fb9043b2bacdf33de29
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02831/task-02831.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:092df0d8fa2871fde3a5dde434d0d087c785278b4d0953f8751a668f278a839d
+size 308359
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02831/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02831/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..ce6713c907c9d22a04bb323b10e47b091ed200c8
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02831/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-02831",
+  "old_fcb_id": "FCB-01699",
+  "task_id": "single_sheet__dag_98d6201706e7",
+  "release_layer": "Core",
+  "prompt_en": "The sheet named \"5-Stock Optimizer\" is a stock-analysis workbook with fields such as Annual Return, Standard Deviation, Yield4, YldYear, LN Blend, LN Index, L_ZLTD, and LLTD_C already laid out. In the Drawdown column, the entries for rows 70 through 321 are blank and need to be filled in. Some of the nearby blank fields in this workbook feed into later fields, so make sure the related row data is in place before completing the Drawdown values in CD70:CD321. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "名为“5-Stock Optimizer”的工作表是一份股票分析工作簿，里面已经设置了 Annual Return、Standard Deviation、Yield4、YldYear、LN Blend、LN Index、L_ZLTD 和 LLTD_C 等字段。Drawdown 列中第 70 行到第 321 行的内容目前为空，需要补全。这个工作簿里有些相邻的空白字段会影响后面的字段，因此请先确保相关行的数据已经补齐，再填写 CD70:CD321 中的 Drawdown 值。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02832/task-02832.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02832/task-02832.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..c8f2a91a7343260ec6d5f8b4fef34f7696a6e21d
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02832/task-02832.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c9dc7a08597813b7a56e3b5f6b2a60b8ea98574e7a2e81ce663576cc0b7c4c0
+size 306933
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02832/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02832/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..ebd07b9879863993ff28f9aabcb2b91327c1cdde
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02832/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-02832",
+  "old_fcb_id": "FCB-01813",
+  "task_id": "single_sheet__dag_e2f6a809b6e8",
+  "release_layer": "Core",
+  "prompt_en": "On the \"5-Stock Optimizer\" sheet, the optimizer table includes fields such as [Person_1]. Annual Return, YEY Custom, CheapHighR26, SomeMo, Standard Deviation, YEYC, YEY, YEY2, Yield4, YldYear, YldDiv, LN Blend, LN Index, L_ZLTD, and LLTD_C. Please complete the missing entries in the t3m column, specifically the cells in CE73:CE321. Some blank fields in this sheet are connected to other blank fields, so make sure any related entries that need to be in place are handled as part of completing t3m. Save the completed workbook as an .xlsx file.",
+  "prompt_cn": "在“5-Stock Optimizer”工作表中，这个优化器表格包含 [Person_1]. Annual Return、YEY Custom、CheapHighR26、SomeMo、Standard Deviation、YEYC、YEY、YEY2、Yield4、YldYear、YldDiv、LN Blend、LN Index、L_ZLTD 和 LLTD_C 等字段。请补全 t3m 列中缺失的内容，具体位置是 CE73:CE321。这个工作表里有些空白字段会影响其他空白字段，所以在完成 t3m 时，也请确保相关需要先补齐的内容已经处理好。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02892/task-02892.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02892/task-02892.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..4788782efb522e2de831485595b931361214a786
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02892/task-02892.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02892/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02892/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..d260fc8f7d06d6683ccc88f6deba1a3ce09692e6
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02892/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-02892",
+  "old_fcb_id": "FCB-07318",
+  "task_id": "single_sheet__dag_df4fbe8c8dce",
+  "release_layer": "Core",
+  "prompt_en": "The Project_Data sheet is a project data sheet with fields like Project Name, Project Manager, Overall Status, and status labels such as Amber. Please complete the blank entries in the Days to Start column on this sheet, covering S10:S351. Some of the project timing fields need to be in place before the related status information is fully meaningful, so keep the entries consistent with the visible project records already on the sheet. When you’re done, please save the completed file as an .xlsx workbook.",
+  "prompt_cn": "Project_Data 工作表是一张项目数据表，里面有 Project Name、Project Manager、Overall Status 等字段，也能看到 Amber 这样的状态标签。请补全该工作表中 Days to Start 列的空白内容，范围是 S10:S351。有些项目时间相关字段需要先补齐，相关状态信息才更完整，因此请让填写内容与表中已有的项目记录保持一致。完成后，请将文件保存为 .xlsx 工作簿。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02982/task-02982.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02982/task-02982.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..69e95680edcf75d3b9aa044f6f223c9fc085b184
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02982/task-02982.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02982/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02982/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..63eb37398b13bbf26b446c356debb23e3d793cae
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-02982/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-02982",
+  "old_fcb_id": "FCB-06903",
+  "task_id": "single_sheet__dag_883671ae3c2e",
+  "release_layer": "Core",
+  "prompt_en": "The sheet named \"general problem \" is laid out as a roll-result table, with columns such as Face ID, prob ??, cumulative??, roll Result, roll result Method 1, and roll result Method 1short. Please fill in the missing entries in the \"roll result Method 1\" column, specifically G2:G400. The result-related columns in this sheet are connected, so the earlier visible and blank entries around the roll results should be in place before the Method 1 column is finalized. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "名为“general problem ”的工作表是一张 roll-result 表，里面有 Face ID、prob ??、cumulative??、roll Result、roll result Method 1 和 roll result Method 1short 等列。请补全“roll result Method 1”列中缺失的内容，具体范围是 G2:G400。这个工作表里的 roll result 相关列彼此有关联，因此在最终填写 Method 1 列之前，周围较早的可见和空白结果项也需要先整理到位。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03069/task-03069.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03069/task-03069.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..48c4a68c0719c1f7cede39f527b2914dac57af0e
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03069/task-03069.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03069/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03069/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..1d0069c802aaf8b2b78bdfa323e649aa966e59e1
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03069/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03069",
+  "old_fcb_id": "FCB-03509",
+  "task_id": "single_sheet__dag_948ae8a22b38",
+  "release_layer": "Core",
+  "prompt_en": "The CTR PROGRESS SPREADSHEET is the EEE forecast progress sheet for the sales pipeline, with contract fields like CTR_NO and CTR_DESCRIPTION, person/resource columns, activity forecast start and finish dates, and dated weekly resource sections. In the RESOURCES REQUIRED column on this sheet, the entries from Q3 through Q224 are blank and need to be filled in for each listed contract/activity row. Some of the supporting activity and weekly resource entries elsewhere in the sheet may need to be in place before this column reads correctly, so complete the visible blanks needed to make the RESOURCES REQUIRED values consistent with the rest of the sheet. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "“CTR PROGRESS SPREADSHEET”工作表是用于销售管线的 EEE 预测进度表，里面有 CTR_NO、CTR_DESCRIPTION 等合同字段，也有人員/资源列、活动预测开始和结束日期，以及按日期列出的每周资源部分。这个工作表中“RESOURCES REQUIRED”列的 Q3 到 Q224 目前为空，需要为每一条合同/活动记录补齐。表中其他活动和每周资源相关的空白内容可能需要先补上，这样“RESOURCES REQUIRED”列才会与其余内容一致。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03109/task-03109.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03109/task-03109.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..6baf9c8d8d8aada420e21558396f3189c9b0e850
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03109/task-03109.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03109/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03109/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..12d39db824caa75cbb51eed5caedb3ebc3996863
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03109/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03109",
+  "old_fcb_id": "FCB-01853",
+  "task_id": "single_sheet__dag_f8b952b5fa57",
+  "release_layer": "Core",
+  "prompt_en": "On the \"Excel Forum Help\" sheet, the rows are laid out around time-related fields like \"Time hh:[Person_1]\", \"Temp\", \"Time hh:mm AM/PM\", \"hh:[Person_1]\", \"hh\", and \"[Person_1]\". Please complete the blank entries in the \"hh:[Person_1]\" column in column G, covering G4:G363. The related time fields on each row need to be in place for the column G entries to make sense, so treat the row data as connected while filling in that column. When you’re done, please save the completed file as an .xlsx workbook.",
+  "prompt_cn": "在“Excel Forum Help”工作表中，各行都是围绕一些与时间相关的字段展开的，比如“Time hh:[Person_1]”、“Temp”、“Time hh:mm AM/PM”、“hh:[Person_1]”、“hh”和“[Person_1]”。请补全 G 列“hh:[Person_1]”中的空白项，范围是 G4:G363。每一行中的相关时间字段需要先保持完整，G 列的内容才有意义，所以填写这一列时请把同一行的数据作为相互关联的内容来看待。完成后，请将补全后的文件保存为 .xlsx 工作簿。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03184/task-03184.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03184/task-03184.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..6d6dd31ea4e07e34e3e336a9422d4014fd40d1b1
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03184/task-03184.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03184/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03184/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..abb68f47c993d4dda75d9347f9e990e8f5b06e4f
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03184/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03184",
+  "old_fcb_id": "FCB-03010",
+  "task_id": "single_sheet__dag_3dc3da20c4f5",
+  "release_layer": "Core",
+  "prompt_en": "The Data sheet is a game-results table organized by Year and Week, with Away Team, Score, Home Team, Winner, Wins, Losses, team-name columns, and a RANK section. The blank entries I need filled are in the Total Score column on the Data sheet, specifically L2:L344. Some of the result-related fields in the sheet feed into the score and ranking area, so make sure the visible game information is used consistently as you complete that Total Score column. Please save the finished workbook as an .xlsx file.",
+  "prompt_cn": "Data 工作表是一张按 Year 和 Week 整理的比赛结果表，里面有 Away Team、Score、Home Team、Winner、Wins、Losses、各球队名称列以及 RANK 区域。需要补全的空白内容在 Data 工作表的 Total Score 列，也就是 L2:L344。表里有些与比赛结果相关的字段会影响得分和排名区域，所以在补全 Total Score 列时，请确保与可见的比赛信息保持一致。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03235/task-03235.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03235/task-03235.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..9535f7e27a4f6238f92155598dd8fb2c5408d5e1
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03235/task-03235.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03235/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03235/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..5b179c516ac53efd3803ec348250cf23dc895b24
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03235/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03235",
+  "old_fcb_id": "FCB-09603",
+  "task_id": "single_sheet__dag_d1bd9a622b9e",
+  "release_layer": "Core",
+  "prompt_en": "The \"PO Increase\" sheet is set up for renewal pricing, with inputs like \"RENEWAL INCREASE PERCENTAGE =\", \"How many years are you activating?\", and \"Does this increase apply to any applicable Cleanings?\" above the line-item table. Please complete the blank entries in the \"RENEWAL Monthly Invoice\" column, specifically column G from G5:G298, for the listed rows with fields such as Sequence, Status, Coverage Type, Description, Item #, Equip Type, Serial #, Start, and End. Some other blank line-item fields in the sheet feed into these renewal invoice entries, so those need to be in place as part of finalizing column G. When finished, please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "“PO Increase”工作表用于续约价格信息，上方有“RENEWAL INCREASE PERCENTAGE =”、“How many years are you activating?”以及“Does this increase apply to any applicable Cleanings?”等输入项，下面是明细表。请补全“RENEWAL Monthly Invoice”这一列中的空白内容，也就是 G 列的 G5:G298，对应的明细行包含 Sequence、Status、Coverage Type、Description、Item #、Equip Type、Serial #、Start 和 End 等字段。工作表中还有一些空白的明细字段会影响这些续约月度发票条目，因此在完善 G 列时也需要让相关明细信息到位。完成后，请将补全后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03236/task-03236.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03236/task-03236.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..394bb45918b58276e3a0850e2eb6c33c227310bf
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03236/task-03236.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03236/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03236/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..e39f43230dd2fb93c76f747eacea1a08107033ff
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03236/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03236",
+  "old_fcb_id": "FCB-09747",
+  "task_id": "single_sheet__dag_eb0a175ff60c",
+  "release_layer": "Core",
+  "prompt_en": "Sheet1 is a data-entry report organized around ARTICLE and QTY details, with matching-related columns such as first match, found DN, DN, ITEM, and the final matching items/articles found field. Please fill in the blank cells in the matching items/articles found column, T2:T367. Some of the matching fields in the sheet feed into this final column, so make sure the related blank entries are in place as needed before completing it. Save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Sheet1 是一张围绕 ARTICLE 和 QTY 信息整理的数据录入报表，里面还有 first match、found DN、DN、ITEM 以及最终的 matching items/articles found 等匹配相关字段。请填写 matching items/articles found 列中的空白单元格，即 T2:T367。表中的一些匹配字段会影响这个最终列，因此在完成该列前，请根据需要先把相关空白内容补齐。完成后请将工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03237/task-03237.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03237/task-03237.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..6ef6d648f78f9fa686ac9a25254b70aa296bd193
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03237/task-03237.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03237/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03237/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..6db11fb1c8e0925212b56b72ff571aa8b19e8e3a
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03237/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03237",
+  "old_fcb_id": "FCB-04032",
+  "task_id": "single_sheet__dag_f58b668d3b91",
+  "release_layer": "Core",
+  "prompt_en": "The Mortgage sheet is laid out for a mortgage payoff review, with visible labels such as Payoff Options:, Interest, #1, #2, and 27 years & 2 months. For the quarterly budget review, please fill in the missing entries in the Balance column on the Mortgage sheet, specifically K11:K335. Some of the monthly entries elsewhere on the sheet may need to be in place before the Balance column reads correctly, so keep the sheet consistent with the visible mortgage schedule. When you’re done, save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Mortgage 工作表是用于查看房贷还款方案的，表中可以看到 Payoff Options:、Interest、#1、#2 和 27 years & 2 months 等标签。为了季度预算 review，请把 Mortgage 工作表中 Balance 列缺失的内容补齐，位置是 K11:K335。表里其他按月列出的项目可能需要先保持完整，Balance 列才会对应得上，所以请让整张房贷明细表与已显示的内容保持一致。完成后，请将补好的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03249/task-03249.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03249/task-03249.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..e5c5a5f49c2b5ca099e1c2d58f601e83733c3348
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03249/task-03249.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03249/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03249/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..0351e8c3fbf66c5a78e7a9376932ac6fbc1693ce
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03249/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03249",
+  "old_fcb_id": "FCB-10548",
+  "task_id": "single_sheet__dag_5be1466b7eb7",
+  "release_layer": "Core",
+  "prompt_en": "The sign-change review on Sheet1 is laid out with columns such as TrueCount, SignChangeAddress, value @ change of sign, value above, and value below, and the main change of sign column still needs to be completed. Please fill in the blank entries in the change of sign column on Sheet1, covering C3:C248. Some of the surrounding blank fields are related to this review, so those supporting entries should be in place where needed for the change of sign column to make sense. Save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Sheet1 上的符号变化检查表包含 TrueCount、SignChangeAddress、value @ change of sign、value above 和 value below 等列，其中主要的 change of sign 列还需要补全。请填写 Sheet1 中 change of sign 列的空白项，范围是 C3:C248。周围有些空白字段也和这项检查相关，因此在需要时应先把这些支撑内容补齐，这样 change of sign 列才完整合理。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03285/task-03285.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03285/task-03285.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..851a0d0bb7a85d293e4af3bcb84d6a69f177a48e
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03285/task-03285.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03285/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03285/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..80862160a34726a4b3f177bec1aed862ac6f4285
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03285/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03285",
+  "old_fcb_id": "FCB-09670",
+  "task_id": "single_sheet__dag_dbb1f0dddc47",
+  "release_layer": "Core",
+  "prompt_en": "The \"material\" sheet is an invoice-style finance workbook with visible fields like INVOICE, DATE:, INVOICE #, FOR:, To:, PER C/M/E, COST EACH, and 90%MARGIN, plus material lines such as EMT & FITTINGS and PER/FT. For the fiscal year close-out review, please complete the blank entries in the 90%MARGIN column on the material sheet, specifically I3:I54, I57:I63, I66:I113, I116:I147, I150:I172, I175:I192, I195:I279, I282:I303, I306:I367, I370:I377, I380:I382, I385:I388, I391:I398, and I401:I409. Some of the other blank cost-related fields in the sheet naturally need to be in place before the 90%MARGIN entries are meaningful, so make sure the sheet is completed consistently. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "“material”工作表是一份发票样式的财务工作簿，里面能看到 INVOICE、DATE:、INVOICE #、FOR:、To:、PER C/M/E、COST EACH、90%MARGIN 等字段，也有 EMT & FITTINGS、PER/FT 这类材料行。为了做财政年度收尾复核，请把 material 工作表中 90%MARGIN 列的空白项补完整，位置包括 I3:I54、I57:I63、I66:I113、I116:I147、I150:I172、I175:I192、I195:I279、I282:I303、I306:I367、I370:I377、I380:I382、I385:I388、I391:I398 和 I401:I409。表里其他一些与成本相关的空白字段需要先保持完整，90%MARGIN 这一列才有意义，所以请确保整张表填写一致。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03286/task-03286.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03286/task-03286.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..0b5fc0c0b27583dd62306a10ae588957f4e5c782
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03286/task-03286.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90c274a3878b63faeded8a77d2a5097fd70ecd08184e143172392cb6c08ff5fc
+size 198810
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03286/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03286/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..1795f40b3c64f37296b2c7dd399ffdaf53cb0ce9
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03286/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03286",
+  "old_fcb_id": "FCB-04516",
+  "task_id": "single_sheet__dag_0b5179be8b11",
+  "release_layer": "Core",
+  "prompt_en": "Sheet1 is a player data table with columns like Rk, Player, Age, Pos, Date, and Tm, and it also includes fantasy-point related fields such as 40 FD PTS. Please finish the blank entries in the 40 FD PTS column on Sheet1, specifically BD2:BD261. A few related blank columns in the sheet may need to be filled in first so the 40 FD PTS column is complete and consistent with the rest of the table. When you’re done, please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Sheet1 是一张球员数据表，里面有 Rk、Player、Age、Pos、Date、Tm 等列，也包含 40 FD PTS 这类与 fantasy points 相关的字段。请补全 Sheet1 中 40 FD PTS 列的空白项，范围是 BD2:BD261。表里还有一些相关的空白列可能需要先补好，这样 40 FD PTS 列才能和整张表保持完整一致。完成后，请将补好的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03287/task-03287.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03287/task-03287.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..80b5f7b9f477bc98b9da51274e258310c21b0c52
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03287/task-03287.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f429aec3dcb4eecae2913d0eccb3512c818dea40d73d21f835ae792cedb6f54
+size 201780
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03287/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03287/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..de333bd985d49883febd3f7b21e2c1e143f3f8d5
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03287/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03287",
+  "old_fcb_id": "FCB-04551",
+  "task_id": "single_sheet__dag_11e127ad19f1",
+  "release_layer": "Core",
+  "prompt_en": "Sheet1 is a player-level data table with game details like Rk, Player, Age, Pos, Date, and Tm, along with fantasy-style point fields including the 15 FD PTS column. Please complete the blank values in the 15 FD PTS column on Sheet1, specifically AY2:AY261. Some other blank columns in the row-level data may need to be in place first so the 15 FD PTS entries fit consistently with the rest of the table. When you’re done, please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Sheet1 是一张按球员记录的明细表，里面有 Rk、Player、Age、Pos、Date、Tm 等比赛信息，也包含类似 fantasy points 的字段，包括 15 FD PTS 列。请补全 Sheet1 中 15 FD PTS 列的空白值，位置是 AY2:AY261。表里其他一些空白列可能需要先补好，这样 15 FD PTS 的内容才能和每一行其余数据保持一致。完成后，请将补全后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03288/task-03288.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03288/task-03288.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..cf055e190560209c128daf46ad95f75ad9e58f95
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03288/task-03288.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc796a832253be56a14a1faf46d3c1f0babea67a7efee68654af199f25172a1c
+size 198541
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03288/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03288/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..d0c2f27fa7bb65515c58bef5f10de57597761585
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03288/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03288",
+  "old_fcb_id": "FCB-04565",
+  "task_id": "single_sheet__dag_1482578586cc",
+  "release_layer": "Core",
+  "prompt_en": "Sheet1 is a player game-log style data table with fields such as Rk, Player, Age, Pos, Date, and Tm, and the FanDuel points section includes a column labeled 50 FD PTS. Please complete the blank entries in the 50 FD PTS column, located in BF2:BF261. Some related scoring columns in the sheet feed into this section naturally, so make sure the supporting blanks are in place where needed before finalizing the 50 FD PTS entries. Save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Sheet1 是一张按球员和比赛日期整理的数据分析表，包含 Rk、Player、Age、Pos、Date、Tm 等字段，FanDuel 得分相关区域里有一列标为 50 FD PTS。请补全 50 FD PTS 列中的空白项，位置是 BF2:BF261。表中有些相关得分列会自然影响到这一部分，所以在完成 50 FD PTS 之前，请确保需要用到的支持性空白项也已补好。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03309/task-03309.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03309/task-03309.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..39911ba55cd2238ddef4efae20907a7dfec018a3
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03309/task-03309.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03309/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03309/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..155c823b7a23ca5201cbb87767162ad953088595
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03309/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03309",
+  "old_fcb_id": "FCB-05098",
+  "task_id": "single_sheet__dag_6cdcc7bc960e",
+  "release_layer": "Core",
+  "prompt_en": "Sheet1 is a report-style data table with service combination columns, and the column labeled “SEI + CPS + WRAP * <=240 units 2 weeks” is still missing its entries for the listed rows. Please fill in the blank values in that column for rows 3 through 300, located in Q3:Q300. Some nearby blank fields in the sheet may need to be in place first so the entries in this column make sense with the rest of the table. When you’re done, save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Sheet1 是一张报表式的数据表，里面有按服务组合划分的列，其中标题为“SEI + CPS + WRAP * <=240 units 2 weeks”的这一列还缺少对应行的内容。请填写该列第 3 行到第 300 行的空白值，位置是 Q3:Q300。表中附近有些空白字段可能也需要先补齐，这样这一列的内容才能和整张表保持一致。完成后，请将补好的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03344/task-03344.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03344/task-03344.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..8e6c1d232dbb0ee1fc9c7eab26d9222d8c4090c6
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03344/task-03344.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4b21ed916e923dea0e22d83bb9c81c882ed5afb0a1a106c55fe43a97b501d84
+size 204206
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03344/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03344/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..32a1a559201befd7b2349d06c4287e19f22e533d
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03344/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03344",
+  "old_fcb_id": "FCB-05970",
+  "task_id": "single_sheet__dag_f6414497ac23",
+  "release_layer": "Core",
+  "prompt_en": "The sheet named \"Transformed by JSON-CSV.CO\" is a booking-style export with fields such as userPnumHash, AllBookingSources, date, appointmentDate, type, amount, category, and isFamily. Please complete the blank entries in the x.differenceDays column, located in column U from U2 through U538. Some blank fields in the sheet may need to be in place before the x.differenceDays column is fully meaningful, so make sure the row-level entries are completed consistently with the visible data. When you’re done, save the completed workbook as an .xlsx file.",
+  "prompt_cn": "名为“Transformed by JSON-CSV.CO”的工作表是一份类似预约记录的导出表，包含 userPnumHash、AllBookingSources、date、appointmentDate、type、amount、category 和 isFamily 等字段。请补全 x.differenceDays 列中的空白内容，位置在 U 列的 U2 到 U538。表中有些空白字段可能需要先补好，x.differenceDays 列的内容才会完整有意义，所以请确保每一行的内容与可见数据保持一致。完成后，请将补全后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03388/task-03388.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03388/task-03388.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..4a50d65a081f50dd90d60cf553b6d3feeccdc0c1
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03388/task-03388.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03388/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03388/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..b3780168106f130189dfdc1d0fba719dadcb919b
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03388/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03388",
+  "old_fcb_id": "FCB-05345",
+  "task_id": "single_sheet__dag_90950c0ebcd3",
+  "release_layer": "Core",
+  "prompt_en": "The Data sheet is an engineering unit listing organized by Year, ZoneDescription, PlantName, PlantID, UnitID, PrimeMoverGroup, FuelType, UnitSize, CapNotonMaint, CapMul12, LongRunAvg, and VarCost12. For this technical data compilation, please fill in the missing entries in the DispatchCost column on the Data sheet, specifically cells U2:U424. Some of the unit and cost-related fields in the sheet naturally support the DispatchCost entries, so those surrounding values should be in place for the DispatchCost column to be completed consistently. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Data 工作表是一份工程机组清单，按 Year、ZoneDescription、PlantName、PlantID、UnitID、PrimeMoverGroup、FuelType、UnitSize、CapNotonMaint、CapMul12、LongRunAvg 和 VarCost12 等字段整理。请在这份技术数据汇编中补全 Data 工作表里 DispatchCost 列的缺失内容，具体范围是 U2:U424。表中的一些机组和成本相关字段会自然支撑 DispatchCost 的填写，因此这些周边数据需要保持一致，DispatchCost 列才能完整。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03389/task-03389.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03389/task-03389.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..4b3811a13462732ae06aedbcc3f1cdc3fdaff8fc
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03389/task-03389.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:962e8bdea3465339469dc59b3f88d320836482e729f8cdc3b1a7ca92ef77233d
+size 208562
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03389/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03389/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..cf16e9e15546be32e7e63a46820200e4dbaa1863
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03389/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03389",
+  "old_fcb_id": "FCB-10730",
+  "task_id": "single_sheet__dag_8e47f3824f6d",
+  "release_layer": "Core",
+  "prompt_en": "The Injury Condensed sheet is part of an ESPN/NFL draft and player ranking workbook, with visible labels such as Saquon Barkley, Christian McCaffrey, Rank, WSID, Round, and Pick. For report finalization, please fill in the blank entries in the Rowname column on the Injury Condensed sheet, covering B3:B661. Some related blank fields in the workbook may need to be in place so the Rowname entries fit the surrounding table. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "“Injury Condensed” 工作表属于一个 ESPN/NFL 选秀和球员排名工作簿，表中可以看到 Saquon Barkley、Christian McCaffrey、Rank、WSID、Round 和 Pick 等标签。为了完成这份报表，请填写 “Injury Condensed” 工作表中 Rowname 列的空白项，范围是 B3:B661。工作簿里一些相关的空白字段可能也需要先补齐，这样 Rowname 的内容才能和周围表格对应起来。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03463/task-03463.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03463/task-03463.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..301421b8838529e024a3f8dc39c6997d482e8ad7
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03463/task-03463.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f49ef0ce18b16d6534706c44240ee465fcc5205d1a0f3ba1bc3f8f3e6f2bab28
+size 122183
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03463/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03463/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..49f4e1ed14701c83f3c6bfd0126a6fc8545a921d
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03463/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03463",
+  "old_fcb_id": "FCB-08900",
+  "task_id": "single_sheet__dag_59df1230e608",
+  "release_layer": "Core",
+  "prompt_en": "The Reference sheet in this sales workbook supports the Weekly Macro Tracker - Meal Planner setup, with visible reference fields like ID and Type already laid out. Please fill in the missing entries in the Combination Field column on the Reference sheet, specifically E3:E500, using the visible reference data and labels in the workbook. Some reference fields naturally feed into other parts of the workbook, so the supporting entries should be in place before the Combination Field is finalized. Please save the completed file as an .xlsx workbook.",
+  "prompt_cn": "这个销售工作簿里的 Reference 工作表用于支持 Weekly Macro Tracker - Meal Planner 的设置，表中已经能看到 ID、Type 等参考字段。请把 Reference 工作表中 Combination Field 列缺失的内容补齐，位置是 E3:E500，并依据工作簿中可见的参考数据和标签来填写。有些参考字段会自然关联到工作簿中的其他部分，所以在最终确认 Combination Field 之前，相关的基础内容也应保持完整。请将完成后的文件保存为 .xlsx 工作簿。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03464/task-03464.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03464/task-03464.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..e6483aa2729e75fdaf65a4de554390bca3f01835
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03464/task-03464.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03464/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03464/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..22a01179d210a2e0d302e701ba4468fa77974dd3
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03464/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03464",
+  "old_fcb_id": "FCB-03566",
+  "task_id": "single_sheet__dag_9f7b0f0e560d",
+  "release_layer": "Core",
+  "prompt_en": "The sheet named \"upbound data\" is part of a river lock barge report, with labels such as \"Figure 11: Upbound Empty Barges,\" \"Mississippi River Locks 27, 1, 52,\" and columns for Upbound and Downbound movements. Please fill in the blank entries in the \"Upbound Empty --% change from last week\" column, specifically cells R8:R457. Some of the weekly movement fields in the sheet feed into this percentage-change column, so make sure the related barge entries are in place as needed while completing it. When you’re done, save the completed file as an .xlsx workbook.",
+  "prompt_cn": "名为“upbound data”的工作表是一份河道船闸驳船报告，其中可以看到“Figure 11: Upbound Empty Barges”、“Mississippi River Locks 27, 1, 52”，以及 Upbound 和 Downbound 相关列。请补全“Upbound Empty --% change from last week”这一列中的空白项，具体范围是 R8:R457。表内有些每周通行数据会影响这一百分比变化列的填写，因此在补全时也请确保相关的驳船数据已按需要填好。完成后，请将文件保存为 .xlsx 工作簿。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03465/task-03465.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03465/task-03465.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..c1a7a41b53bc874885d56966d4c550092efc4ac9
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03465/task-03465.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03465/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03465/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..8d73b3edaad0cdcbf8dc139e7b785949ad16476d
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03465/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03465",
+  "old_fcb_id": "FCB-03909",
+  "task_id": "single_sheet__dag_ddb294d1cfe0",
+  "release_layer": "Core",
+  "prompt_en": "The “upbound data” sheet is part of the Figure 11 report for Upbound Empty Barges on the Mississippi River, with lock-related columns such as Upbound 27, Downbound 27, Upbound 1, and Downbound 1 already laid out. Please complete the blank entries in the “Upbound Empty --% change from last week” column, located in column R, rows 8 through 439. Some of the nearby weekly barge figures need to be in place before this percent-change column makes sense, so use the visible data on the sheet to finish the missing values consistently. When you’re done, please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "“upbound data”工作表是密西西比河 Figure 11 上行空驳报告的一部分，表中已经列出了与船闸相关的列，例如 Upbound 27、Downbound 27、Upbound 1 和 Downbound 1。请补全“Upbound Empty --% change from last week”这一列中的空白项，位置在 R 列第 8 行到第 439 行。周边的一些每周驳船数据需要先齐全，这个百分比变化列才有意义，所以请根据工作表中可见的数据把缺失值补完整并保持一致。完成后，请将补好的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03466/task-03466.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03466/task-03466.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..b7f72523d20bd722594ba66c10aa06238ed11782
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03466/task-03466.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03466/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03466/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..c8813a88dc00c277b929c16d01d67c3c8f79cc8d
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03466/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03466",
+  "old_fcb_id": "FCB-09690",
+  "task_id": "single_sheet__dag_de7aa1b01c63",
+  "release_layer": "Core",
+  "prompt_en": "The weekday sheet is laid out as a full-year calendar table, with Date, WeekName, Number of Week Day, and Qtr columns side by side. The Qtr column still needs to be completed: please fill in the blank quarter entries in D2:D366 on the weekday sheet. The row-level date and weekday information needs to be in place for the Qtr entries to make sense alongside the rest of the table. Please save the completed file as an .xlsx workbook.",
+  "prompt_cn": "weekday 工作表是一张按全年日期整理的日历表，旁边依次有 Date、WeekName、Number of Week Day 和 Qtr 等列。现在需要补全 Qtr 列：请填写 weekday 工作表中 D2:D366 的空白季度条目。每一行的日期和星期信息需要与 Qtr 条目对应起来，这样整张表才完整一致。请将完成后的文件保存为 .xlsx 工作簿。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03532/task-03532.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03532/task-03532.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..088cace6302509ff2df54aa37673c0b9816b99ac
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03532/task-03532.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03532/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03532/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..953e70486e2493db5e9b6105826be2e621cedcee
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03532/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03532",
+  "old_fcb_id": "FCB-03413",
+  "task_id": "single_sheet__dag_80b3c0c66961",
+  "release_layer": "Core",
+  "prompt_en": "The “upbound data” sheet is the Figure 11 table for upbound empty barges at Mississippi River Locks 27, 1, and 52, with columns such as Upbound 27, Downbound 27, Upbound 1, and Downbound 1. Please fill in the blank entries in the “Change in number of upbound empty barges from previous week” column, specifically S8:S410. Some of the weekly barge fields on this sheet feed naturally into that change field, so the relevant figures should be in place before the change entries are finalized. Save the completed workbook as an .xlsx file.",
+  "prompt_cn": "“upbound data” 工作表是关于密西西比河 27、1 和 52 号船闸上行空驳船的 Figure 11 表格，其中包含 Upbound 27、Downbound 27、Upbound 1、Downbound 1 等列。请填写 “Change in number of upbound empty barges from previous week” 这一列中的空白项，具体位置是 S8:S410。这个工作表里有些每周驳船数据字段会自然关联到这个变化字段，因此相关数据应先补齐后，这一列的内容才完整。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03533/task-03533.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03533/task-03533.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..22ce5681d1cec43f969079b1052d90e730120abc
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03533/task-03533.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03533/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03533/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..db2a8503ee5c3501cea52d1e556e091758ec1451
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03533/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03533",
+  "old_fcb_id": "FCB-05510",
+  "task_id": "single_sheet__dag_ad28dbc2cc26",
+  "release_layer": "Core",
+  "prompt_en": "The “upbound data” sheet is a weekly barge movement table for “Figure 11: Upbound Empty Barges,” with Mississippi River Locks 27, 1, and 52 shown alongside fields like Upbound 27, Downbound 27, Upbound 1, and Downbound 1. Please complete the entries in the “Change in number of upbound empty barges from previous week” column, specifically S8:S439. Some of the weekly barge-count fields need to be in place before that change column is meaningful, so use the visible data on the sheet to finish the missing entries consistently. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "“upbound data”工作表是一张每周驳船通行数据表，对应“Figure 11: Upbound Empty Barges”，其中列出了 Mississippi River Locks 27、1、52，以及 Upbound 27、Downbound 27、Upbound 1、Downbound 1 等字段。请补全“Change in number of upbound empty barges from previous week”这一列中的内容，位置是 S8:S439。有些每周驳船数量字段需要先补齐后，这个变化列才有意义，所以请根据表中可见数据把缺失内容填写完整并保持一致。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03534/task-03534.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03534/task-03534.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..8f77be799a49f382684b472a7565489738828a68
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03534/task-03534.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c5908bf42138374fbfb9c20a048ef29e7e24adbacbd3e4beadf0785689e12701
+size 107484
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03534/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03534/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..0be0af776b993ad697e73c306382f30d7f9b48b1
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03534/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03534",
+  "old_fcb_id": "FCB-09417",
+  "task_id": "single_sheet__dag_b21ba12376cb",
+  "release_layer": "Core",
+  "prompt_en": "The “upbound data” sheet is a barge movement report tied to Figure 11, with visible fields for Mississippi River Locks 27, 1, and 52, including Upbound 27, Downbound 27, Upbound 1, and Downbound 1. Please complete the blank entries in the “Change in number of upbound empty barges from previous week” column, located in column S for rows 8 through 492. Some of the weekly barge movement columns need to be in place before this change column is meaningful, so make sure the related visible data on the sheet is considered as you fill it in. Save the completed workbook as an .xlsx file.",
+  "prompt_cn": "“upbound data”工作表是一份与 Figure 11 相关的驳船运行数据报表，里面可以看到 Mississippi River Locks 27、1、52，以及 Upbound 27、Downbound 27、Upbound 1、Downbound 1 等字段。请补全“Change in number of upbound empty barges from previous week”这一列中的空白项，位置是 S 列第 8 行到第 492 行。部分每周驳船运行相关列需要先具备完整内容，这一变化列才有意义，因此填写时请结合工作表中可见的相关数据。完成后请将工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03535/task-03535.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03535/task-03535.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..9624c2cc231e5ecdb57b376267274f442c8a500c
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03535/task-03535.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03535/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03535/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..b71893bce538befcf735ce92979b1490d774929d
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03535/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03535",
+  "old_fcb_id": "FCB-05645",
+  "task_id": "single_sheet__dag_c17ee0622d3f",
+  "release_layer": "Core",
+  "prompt_en": "The \"upbound data\" sheet is tracking empty barge movements for Figure 11: Upbound Empty Barges, with Mississippi River lock columns such as Upbound 27, Downbound 27, Upbound 1, and Downbound 1. Please complete the blank entries in the \"Change in number of upbound empty barges from previous week\" column, specifically S8:S457. Some related blank columns on the sheet feed into this change column, so make sure the underlying weekly barge figures are in place wherever they are needed before finalizing column S. When finished, please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "“upbound data”工作表用于记录“Figure 11: Upbound Empty Barges”的空驳船通行数据，表中包含 Mississippi River 船闸相关列，例如 Upbound 27、Downbound 27、Upbound 1 和 Downbound 1。请补全“Change in number of upbound empty barges from previous week”这一列中的空白项，位置是 S8:S457。表中有些相关的空白列会影响这个变化列，因此在最终补全 S 列之前，请先确保需要用到的每周驳船数据都已填好。完成后，请将补全后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03537/task-03537.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03537/task-03537.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..c2dec8df429c7a834129026d2e03ab6cb4d60517
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03537/task-03537.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5dda14ae7d86d2b7d21c2eca405a152f29a134453b2ee8ea454f2897195636a
+size 424190
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03537/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03537/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..ceefe3cf46b625b1f518a609bf6294da527e17a2
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03537/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03537",
+  "old_fcb_id": "FCB-08460",
+  "task_id": "single_sheet__dag_0b94ab09039b",
+  "release_layer": "Core",
+  "prompt_en": "DataOld is the finance sheet tying Store Entity and Account Number details to monthly columns such as JAN, FEB, MAR, APR, MAY, and YEAR END, with account descriptions like Sales-Automotive, Sales-Advertising Program, and Sales-Builders Hardware. For the fiscal year close-out view, please complete the blank entries in the FinalAccount column on DataOld, specifically W2:W306. Some of the account and monthly fields in the workbook feed into the final account information, so those details need to be in place before the FinalAccount column makes sense. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "DataOld 是一张财务工作表，把 Store Entity 和 Account Number 等信息与 JAN、FEB、MAR、APR、MAY、YEAR END 等月份和年末栏位对应起来，里面还有 Sales-Automotive、Sales-Advertising Program、Sales-Builders Hardware 等账户描述。为了完成财政年度收尾视图，请补全 DataOld 表中 FinalAccount 列的空白内容，范围是 W2:W306。工作簿里的部分账户和月份信息会影响最终账户信息，所以这些相关内容需要先对应完整，FinalAccount 列才有意义。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03538/task-03538.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03538/task-03538.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..feab9ea3438bd10e16ab7765e67db6cde7af863a
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03538/task-03538.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03538/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03538/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..eb16e371d6201b51f47b2806c978be427acd1b05
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03538/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03538",
+  "old_fcb_id": "FCB-10966",
+  "task_id": "single_sheet__dag_cf5b3dd3f1fd",
+  "release_layer": "Core",
+  "prompt_en": "Sheet1 is a roster-style table built around Date of Birth, Age, Age limit, and Pay rate, with the Pay rate column still missing for the listed records. Please fill in the blank Pay rate entries in column E, specifically E3:E338. The age-related fields in the sheet sit alongside the pay information, so those entries need to be in place for the Pay rate column to make sense. When you’re done, please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Sheet1 是一张类似人员清单的表，围绕 Date of Birth、Age、Age limit 和 Pay rate 这些字段整理数据，其中 Pay rate 列还有一批记录没有填好。请补全 E 列中的空白 Pay rate 项，范围是 E3:E338。表里的年龄相关字段和薪资信息是放在一起使用的，因此这些信息需要就位后，Pay rate 列才完整清楚。完成后，请将补全后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03605/task-03605.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03605/task-03605.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..6eab25e425776deb8cea1967dcf62483ad1dc430
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03605/task-03605.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03605/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03605/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..bbfd46ca26cb9226dc642d17904c8818627b6aa1
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03605/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03605",
+  "old_fcb_id": "FCB-03190",
+  "task_id": "single_sheet__dag_5c8ab24b57f6",
+  "release_layer": "Core",
+  "prompt_en": "The Suppliers sheet is a finance tracker for supplier invoices, with sections and headers like Cash Flow, Outgoing, Incoming, Total Unpaid, 30+ Days, Payment Status, Invoice Number, Job Number, Supplier, and Payment Method. For the fiscal year close-out review, please complete the blank entries in the GROSS (Net + VAT) column on the Suppliers sheet, covering I2:I500. Some of the invoice detail columns in this sheet need to be in place before the gross figures make sense, so use the visible supplier invoice information already laid out in the workbook to fill the missing gross amounts. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Suppliers 工作表是一个供应商发票的财务跟踪表，里面能看到 Cash Flow、Outgoing、Incoming、Total Unpaid、30+ Days、Payment Status、Invoice Number、Job Number、Supplier 和 Payment Method 等区块和字段。为了财政年度收尾复核，请补全 Suppliers 工作表中 GROSS (Net + VAT) 列的空白内容，范围是 I2:I500。这个表里有些发票明细列需要先完整，毛额栏位才有意义，所以请根据工作簿中已经显示的供应商发票信息，把缺失的毛额填写完整。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03606/task-03606.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03606/task-03606.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..9d55ce8278cf96da001356ac632075a529486f18
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03606/task-03606.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03606/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03606/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..6e6998a4e5466d40eaa1c77a6a8ac442f1e5e9f6
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03606/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03606",
+  "old_fcb_id": "FCB-04071",
+  "task_id": "single_sheet__dag_fd0989d2e7c4",
+  "release_layer": "Core",
+  "prompt_en": "Sheet2 is the finance working sheet that connects the DATA bank transaction columns like Bank, Date, Description, and Amount with the Public Holidays area and the START/FINAL date section. In the date run on Sheet2, the Day of week column is blank for the daily entries.\n\nPlease fill in the missing Day of week values in Sheet2, column B, for the range B5:B369, matching each row’s visible date context. Those weekday entries need to be in place before the related public-holiday and bank-day fields on the sheet make sense.\n\nPlease save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Sheet2 是这个财务工作表，用来把 DATA 中的银行交易字段（如 Bank、Date、Description、Amount）和 Public Holidays 区域以及 START/FINAL 日期部分联系在一起。Sheet2 的日期序列里，Day of week 这一列在每日记录中还是空的。\n\n请在 Sheet2 的 B 列补全缺失的 Day of week，范围是 B5:B369，并与每一行可见的日期内容对应起来。这些星期信息需要先填好，表中相关的公共假日和银行日字段才有意义。\n\n请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03607/task-03607.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03607/task-03607.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..27858a317f2cfbd310c9374179cf7175bb8da515
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03607/task-03607.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03607/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03607/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..c85d6163f055b19d36d166c2daf71b81440bbaf7
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03607/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03607",
+  "old_fcb_id": "FCB-04078",
+  "task_id": "single_sheet__dag_ff06fa1cd587",
+  "release_layer": "Core",
+  "prompt_en": "The sheet \"2011-12 Ch. 41 Districts\" is an education enrollment report listing districts by CDN and DISTRICT NAME, with fields such as 2010 DPV, WADA42 (HB3646), ENROLLMENT, and WADA TO ENROLL RATIO. Please fill in the blank WADA41 entries in column I for the district rows, specifically I2:I357. Some of the related blank district figures in the sheet need to be in place before the WADA41 column is fully meaningful, so make sure the completed entries fit with the rest of the visible district data. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "工作表“2011-12 Ch. 41 Districts”是一份教育领域的入学相关报表，按 CDN 和 DISTRICT NAME 列出各个学区，并包含 2010 DPV、WADA42 (HB3646)、ENROLLMENT、WADA TO ENROLL RATIO 等字段。请补全学区行中 I 列的 WADA41 空白项，范围是 I2:I357。表中一些相关的学区空白数据需要先完整，WADA41 列的内容才会与整张表的数据对应起来，所以请确保填入的内容与可见的学区数据一致。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03666/task-03666.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03666/task-03666.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..598adf4c622f53d8677a8e91b892c936da8651c1
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03666/task-03666.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41ea4bfc487c5bfcf702828d69d3435c01b180631f2d875adfda2fda17f74089
+size 184979
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03666/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03666/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..1773cafa472289930e66550b668b55032965b8be
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03666/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03666",
+  "old_fcb_id": "FCB-05713",
+  "task_id": "single_sheet__dag_cc5d4e2e8d8d",
+  "release_layer": "Core",
+  "prompt_en": "Sheet1 is laid out like a hotel pricing and occupancy report, with columns such as Price Room 1 through Price Room 6, RPD, H-Occ% LY TODAY, Occ% [Person_1], SPO LY TODAY, RTPD, and room categories like Ceiling, DLX, PREM, Junior S., 1 Bdr Dlx S., and 1 Bdr Prm S. The H-Occ% column in column K is still blank for the main body of the report, so please fill in the H-Occ% entries in Sheet1, K3:K498. Some of the other blank fields in the sheet are part of the same report flow, so those related entries may need to be in place before the H-Occ% column is complete. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Sheet1 看起来是一张酒店价格和入住率报表，里面有 Price Room 1 到 Price Room 6、RPD、H-Occ% LY TODAY、Occ% [Person_1]、SPO LY TODAY、RTPD 等列，也包含 Ceiling、DLX、PREM、Junior S.、1 Bdr Dlx S.、1 Bdr Prm S. 这些房型类别。现在 K 列的 H-Occ% 在报表主体部分还没有填好，请把 Sheet1 中 K3:K498 的 H-Occ% 项目补完整。表里还有一些其他空白字段属于同一份报表流程，相关内容可能需要先补齐，H-Occ% 列才会完整。完成后请将工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03667/task-03667.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03667/task-03667.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..6c3932f19818d9938b7dc49eef57adfde7faf5ee
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03667/task-03667.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03667/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03667/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..34a4116837ce497e8e0419d362514dfc5abe953d
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03667/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03667",
+  "old_fcb_id": "FCB-04803",
+  "task_id": "single_sheet__dag_3cb5ce35e6e5",
+  "release_layer": "Core",
+  "prompt_en": "The “2010” sheet is an academic records-style calendar with columns for Year, Month, Date, Day of Week, Work Day, and Holidays. The Date column is blank for the main daily entries, so please fill in the missing Date values in E5:E369. Those date entries need to be in place before the related Day of Week, Work Day, and Holidays fields can be completed reliably. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "“2010”工作表是一张用于学籍或教学记录的日历表，包含 Year、Month、Date、Day of Week、Work Day 和 Holidays 等列。当前主要每日记录中的 Date 列为空，请补全 E5:E369 中缺失的 Date 值。这些日期信息需要先填写好，后面的 Day of Week、Work Day 和 Holidays 字段才好继续完善。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03668/task-03668.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03668/task-03668.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..dc6ed778258536cd0acbce3c777d4c90ea4b115a
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03668/task-03668.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03668/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03668/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..d66ce66b9aee8f6f2084c36174011f5caef4f767
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03668/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03668",
+  "old_fcb_id": "FCB-08978",
+  "task_id": "single_sheet__dag_663809838ba4",
+  "release_layer": "Core",
+  "prompt_en": "The Data sheet is a Basic Data log with calendar information alongside columns like Date, Extra, Water mtr, and Solar prod, but the Day column at the far left is blank for the main set of entries. Please fill in the missing Day values in Sheet \"Data\", range A4:A795, so the day labels are complete for each visible record. Some of the other columns are meant to sit alongside these calendar entries, so the Day values should be in place before the sheet is treated as finished. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "“Data”工作表是一张“Basic Data”记录表，里面有 Date、Extra、Water mtr、Solar prod 等列，但最左侧的 Day 列在主要记录区域还是空的。请把“Data”工作表 A4:A795 范围内缺失的 Day 值补全，让每条可见记录都有对应的日期标签。其他一些列是和这些日历记录配套使用的，所以在确认整张表完成之前，需要先把 Day 列填好。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03669/task-03669.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03669/task-03669.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..e41aab92496fee6ab40529445e68d26a0a858064
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03669/task-03669.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03669/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03669/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..244ba64bbd9940b5be31fca87fd98cad3272c5de
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-03669/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-03669",
+  "old_fcb_id": "FCB-04060",
+  "task_id": "single_sheet__dag_fab134a480a5",
+  "release_layer": "Core",
+  "prompt_en": "The sheet named \"2010\" is set up like an academic records calendar, with columns labeled Year, Month, Date, Day of Week, Work Day, and Holidays. For the enrollment report update, please fill in the missing Month entries in the Month column on this sheet, covering D5:D369. Those month entries need to be in place before the related calendar fields are meaningful. When you’re done, please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "名为“2010”的工作表看起来是用于学籍记录的日历表，列标题包括 Year、Month、Date、Day of Week、Work Day 和 Holidays。为了更新 enrollment report，请把这个工作表中 Month 列缺失的月份内容补全，范围是 D5:D369。这些月份内容需要先填写好，相关的日历字段才有意义。完成后，请将补全后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05496/task-05496.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05496/task-05496.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..5010557879019eaefccec88bb29e5f7bfe579fc1
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05496/task-05496.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f95500a3292407dad3284e9769e520aaaa5dee3bd578b63b08b3c7751233f5c5
+size 612218
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05496/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05496/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..c8d4b3f6ac48ff9418d8cf17f7506ba9fa87d295
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05496/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-05496",
+  "old_fcb_id": "FCB-05210",
+  "task_id": "single_sheet__dag_7b836446edbc",
+  "release_layer": "Core",
+  "prompt_en": "Sheet10 is a finance worksheet centered on Price Performance for ASX tickers, with visible fields such as TICKER, Industry, Company, Price, Capital Structure, Historical Financials, Balance Sheet, and interest-related items like 10 Year bonds. For the quarterly budget review / fiscal year close-out, please complete the blank entries in the Cumulative Profit column on Sheet10, specifically H3:H576. If any related blank finance fields on the sheet are needed for the Cumulative Profit entries to make sense, make sure those are filled in first. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Sheet10 是一张围绕 ASX 股票 Price Performance 的财务工作表，能看到 TICKER、Industry、Company、Price、Capital Structure、Historical Financials、Balance Sheet，以及 10 Year bonds 等与利息相关的字段。为了季度预算复核 / 财年收尾，请补全 Sheet10 中 Cumulative Profit 列的空白项，具体范围是 H3:H576。如果表中还有与这些 Cumulative Profit 项目相关的空白财务字段需要先补齐，先把它们填好。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05893/task-05893.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05893/task-05893.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..273808c21e91d0e2031bcb7d6d7e2fcf8ea011c0
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05893/task-05893.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05893/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05893/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..92e549cc2c972537a966d15a3244ac655d5fa6f0
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05893/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-05893",
+  "old_fcb_id": "FCB-11159",
+  "task_id": "single_sheet__dag_3e509ba15c82",
+  "release_layer": "Core",
+  "prompt_en": "On the “general problem ” sheet, the Face ID rows are organized with the “prob ??”, “cumulative??”, “roll Result”, “roll result Method 1”, and “roll result Method 1short” fields. The short Method 1 result field is still blank for the main run of entries.\n\nPlease fill in the missing values in the “roll result Method 1short” column, specifically H2:H400. The related roll result columns should be in place as needed before the short result entries are finalized. When you’re done, save the completed workbook as an .xlsx file.",
+  "prompt_cn": "在“general problem ”工作表中，Face ID 各行按“prob ??”、“cumulative??”、“roll Result”、“roll result Method 1”和“roll result Method 1short”等字段整理。现在主要条目中的简短 Method 1 结果字段还留空。\n\n请填写“roll result Method 1short”列中的缺失值，具体范围是 H2:H400。相关的 roll result 列需要先补齐到位，这样简短结果条目才能最终完成。完成后，请将工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05962/task-05962.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05962/task-05962.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..fb214899ca5118913fdef12cb19bd43ed2a4d91b
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05962/task-05962.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7f84d37a323f9250d5ab960326f732e98360dfab9fab20c9e16a04fc8a36853e
+size 521117
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05962/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05962/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..dc47daa0cce603fb40bfcca7cf751683b7e51cef
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-05962/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-05962",
+  "old_fcb_id": "FCB-07064",
+  "task_id": "cross_sheet__dag_3988cbb492cc",
+  "release_layer": "Core",
+  "prompt_en": "The “Tax Bill Summary” sheet is part of the FY16 finance workbook, alongside references like “FY 2016 as of February Forecast - Max Revenue” and “FY 2016 as of 2014Tax Bill Changes from February Forecast -”, with location categories such as “MPLS & ST PAUL”, “OTHER METRO”, and “NONMET < 1K”. For the fiscal year close-out review, please fill in the missing entries in the “Aid” column on the “Tax Bill Summary” sheet: AO9, AO11:AO19, and AO21:AO517. The related FY 2016 revenue and tax bill change fields elsewhere in the workbook should be in place before the Aid column is complete. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "“Tax Bill Summary” 工作表是这份 FY16 财务工作簿的一部分，工作簿里还包含 “FY 2016 as of February Forecast - Max Revenue” 和 “FY 2016 as of 2014Tax Bill Changes from February Forecast -” 等内容，并按 “MPLS & ST PAUL”、“OTHER METRO”、“NONMET < 1K” 等地区类别列示。为了财政年度收尾复核，请补全 “Tax Bill Summary” 工作表中 “Aid” 列缺失的内容：AO9、AO11:AO19，以及 AO21:AO517。工作簿中相关的 FY 2016 收入和税单变动字段需要先保持完整，这样 Aid 列才有意义。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_cross",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06116/task-06116.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06116/task-06116.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..694c4ad3ece6c7e0065d351fd8cd40cedf8693b7
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06116/task-06116.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a9867c8651dbfac5d7a64ebdf7ed3c8ce0e7b2904c674c749b93dc24c6709a6
+size 157720
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06116/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06116/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..78c567e48ea98e0cd68ba605be13e880cac32d64
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06116/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-06116",
+  "old_fcb_id": "FCB-05951",
+  "task_id": "single_sheet__dag_f20a73ddf028",
+  "release_layer": "Core",
+  "prompt_en": "On the “Daily Prod. Stats Worksheet,” the Daily Lead QC / DAILY PRODUCTION REPORT tracks project activity by day, with fields like Mulching (km), Slashing (km), Start Survey Total (km), Expected Finish, and Day. Please fill in the missing entries in the “Balance remaining to complete” column, column G, for the range G7:G370. Because this remaining-balance column sits alongside the daily production fields, the related daily figures should be in place first so the completed column is consistent with the rest of the report. Once complete, save the workbook as an .xlsx file.",
+  "prompt_cn": "在“Daily Prod. Stats Worksheet”工作表中，这份 Daily Lead QC / DAILY PRODUCTION REPORT 按天记录项目活动，包含 Mulching (km)、Slashing (km)、Start Survey Total (km)、Expected Finish 和 Day 等字段。请填写“Balance remaining to complete”这一列，也就是 G 列中 G7:G370 范围内的空白内容。由于这一列与日常生产字段并列使用，相关的每日数据应先补齐，这样填写后的剩余完成量列才能与报告其余部分保持一致。完成后，请将工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06133/task-06133.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06133/task-06133.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..2ed889ab5786aa529f103d15f96ba80a413c1e04
Binary files /dev/null and b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06133/task-06133.xlsx differ
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06133/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06133/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..293e3e1f0a8be8e5922e4234594eab5ef9381937
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06133/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-06133",
+  "old_fcb_id": "FCB-03720",
+  "task_id": "single_sheet__dag_b8568a9e4403",
+  "release_layer": "Core",
+  "prompt_en": "The sheet named \"NJ LIST - EXCEL FORUM 10062011\" is an address list with columns like Full Name, Address Line 1, Address Line 2, City, State, and ZIP Code, and the duplicate-tracking area at the right still needs to be completed. Please fill in the blank cells in the column labeled \"count dups\" in column J, from J2 through J300. Make sure the related blank address-list fields that feed into this duplicate-tracking column are in place so the entries in \"count dups\" make sense. When you’re done, save the completed file as an .xlsx workbook.",
+  "prompt_cn": "名为“NJ LIST - EXCEL FORUM 10062011”的工作表是一份地址清单，里面有 Full Name、Address Line 1、Address Line 2、City、State 和 ZIP Code 等列，右侧用于重复项跟踪的部分还需要补全。请填写 J 列中标为“count dups”的空白单元格，范围是 J2 到 J300。请确保与该重复项跟踪列相关的地址清单空白字段也已补齐，这样“count dups”中的内容才完整合理。完成后，请将文件保存为 .xlsx 工作簿。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_single",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06510/task-06510.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06510/task-06510.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..74042efce811bb1a92bdfce585da89e7497dfeda
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06510/task-06510.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2dfe9051ccef617b5c12e2d5a266295b1f6f2522bd7c4245a314144136732af
+size 680414
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06510/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06510/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..7215fb7f5182ae3486da8d8d04d3ebb9ad091294
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06510/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-06510",
+  "old_fcb_id": "FCB-09943",
+  "task_id": "cross_sheet__dag_43041537ee06",
+  "release_layer": "Core",
+  "prompt_en": "The Forecast Data sheet is being used for finance review work around FORECAST entries, Actual Date, Payment Term Variance, and currency-related fields. Please complete the missing values in the GBP column on the Forecast Data sheet, specifically cells P2:P894. Some of the blank fields in the workbook naturally feed into other blank fields, so make sure the supporting forecast and currency information visible across the workbook is in place as needed before finalizing the GBP column. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Forecast Data 工作表用于财务复核，内容涉及 FORECAST 条目、Actual Date、Payment Term Variance 以及与币种相关的字段。请补全 Forecast Data 工作表中 GBP 列的缺失值，具体范围是 P2:P894。工作簿中有些空白字段会自然影响其他空白字段，因此在最终完成 GBP 列之前，请根据需要先确认工作簿中可见的预测和币种相关信息已经补齐。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_cross",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06514/task-06514.xlsx b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06514/task-06514.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..ea8784f19bc166b4f4a54d6920d062ee53ec1bba
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06514/task-06514.xlsx
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce9233fad3273b228860f99ebfb0def89bde21f013e2586097acb6f938160eb7
+size 685121
diff --git a/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06514/task_description.json b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06514/task_description.json
new file mode 100644
index 0000000000000000000000000000000000000000..ef3a8c37f80fcca8d2f403e51d6a32bd73019a2f
--- /dev/null
+++ b/FormulaCascade_Benchmark_eval_200_Release_version/release/FCB-06514/task_description.json
@@ -0,0 +1,13 @@
+{
+  "fcb_id": "FCB-06514",
+  "old_fcb_id": "FCB-10154",
+  "task_id": "cross_sheet__dag_e69a7091f5ef",
+  "release_layer": "Core",
+  "prompt_en": "The Forecast Data sheet is part of a finance workbook used around FORECAST, payment term variance, and actual payment timing, with columns such as Type, Text, CoCd, Co Currency, Co Amount, Doc. Date, Net Date, Actual Date, and weekday labels. In the Forecast Data sheet, please fill in the missing entries in the Paid on weekday column, located in V2:V894. Some of the date-related information in the workbook needs to be in place before the weekday field is meaningful, so use the visible data across the workbook to complete those blanks consistently. Please save the completed workbook as an .xlsx file.",
+  "prompt_cn": "Forecast Data 工作表属于一个财务工作簿，用来记录 FORECAST、付款期限差异以及实际付款时间，表中能看到 Type、Text、CoCd、Co Currency、Co Amount、Doc. Date、Net Date、Actual Date 以及各星期几等字段。请在 Forecast Data 工作表中补全 Paid on weekday 列缺失的内容，位置是 V2:V894。工作簿里有些与日期相关的信息需要先对应完整，这样 weekday 字段才有意义；请根据工作簿中可见的数据，把这些空白内容一致地补齐。请将完成后的工作簿保存为 .xlsx 文件。",
+  "source_release_layer_v13": "Broad",
+  "source_release_layer_label_v13": null,
+  "paper_stratum_v5": "Core",
+  "eval_group": "Core_cross",
+  "release_layer_label": "structural transfer pool"
+}
\ No newline at end of file