Datasets:
task_id stringlengths 10 19 | domain stringclasses 2
values | level int64 1 3 | seed int64 0 2 | instruction stringlengths 267 1.01k | files unknown | oracle unknown | hidden_tests int64 0 29 |
|---|---|---|---|---|---|---|---|
sw_stats_fix-l1 | software engineering | 1 | 0 | `pytest tests/` fails for the `numutil` module in the current directory. Find and fix the bug in numutil.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fix the cause rather than special-casi... | {
"numutil.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle valu... | {
"numutil.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle valu... | 29 |
sw_stats_fix-l1 | software engineering | 1 | 1 | `pytest tests/` fails for the `numutil` module in the current directory. Find and fix the bug in numutil.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fix the cause rather than special-casi... | {
"numutil.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle valu... | {
"numutil.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle valu... | 29 |
sw_stats_fix-l1 | software engineering | 1 | 2 | `pytest tests/` fails for the `stats` module in the current directory. Find and fix the bug in stats.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fix the cause rather than special-casing t... | {
"stats.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) // len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle value... | {
"stats.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle values... | 29 |
sw_stats_fix-l2 | software engineering | 2 | 0 | A user reported a bug in the `stats` module in the current directory (see ISSUE.md); the existing tests in tests/ still pass. Fix stats.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fix the... | {
"stats.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle values... | {
"stats.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle values... | 29 |
sw_stats_fix-l2 | software engineering | 2 | 1 | A user reported a bug in the `numutil` module in the current directory (see ISSUE.md); the existing tests in tests/ still pass. Fix numutil.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fix... | {
"numutil.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) // len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle val... | {
"numutil.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle valu... | 29 |
sw_stats_fix-l2 | software engineering | 2 | 2 | A user reported a bug in the `stats` module in the current directory (see ISSUE.md); the existing tests in tests/ still pass. Fix stats.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fix the... | {
"stats.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle values... | {
"stats.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle values... | 29 |
sw_stats_fix-l3 | software engineering | 3 | 0 | The `stats` module in the current directory has two bugs: `pytest tests/` fails, and ISSUE.md reports another problem. Fix both in stats.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fix th... | {
"stats.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle values... | {
"stats.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle values... | 29 |
sw_stats_fix-l3 | software engineering | 3 | 1 | The `numutil` module in the current directory has two bugs: `pytest tests/` fails, and ISSUE.md reports another problem. Fix both in numutil.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fi... | {
"numutil.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) // len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle val... | {
"numutil.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle valu... | 29 |
sw_stats_fix-l3 | software engineering | 3 | 2 | The `numutil` module in the current directory has two bugs: `pytest tests/` fails, and ISSUE.md reports another problem. Fix both in numutil.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fi... | {
"numutil.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle valu... | {
"numutil.py": "\"\"\"Small descriptive-statistics helpers used by the reporting jobs.\"\"\"\n\n\ndef mean(xs):\n \"\"\"Arithmetic mean of a non-empty list of numbers.\"\"\"\n return sum(xs) / len(xs)\n\n\ndef median(xs):\n \"\"\"Median of a non-empty list; for an even count, the mean of the two middle valu... | 29 |
sw_dates_fix-l1 | software engineering | 1 | 0 | `pytest tests/` fails for the `billing_dates` module in the current directory. Find and fix the bug in billing_dates.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fix the cause rather than ... | {
"billing_dates.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fro... | {
"billing_dates.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fro... | 25 |
sw_dates_fix-l1 | software engineering | 1 | 1 | `pytest tests/` fails for the `calendar_utils` module in the current directory. Find and fix the bug in calendar_utils.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fix the cause rather tha... | {
"calendar_utils.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fr... | {
"calendar_utils.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fr... | 25 |
sw_dates_fix-l1 | software engineering | 1 | 2 | `pytest tests/` fails for the `calendar_utils` module in the current directory. Find and fix the bug in calendar_utils.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fix the cause rather tha... | {
"calendar_utils.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fr... | {
"calendar_utils.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fr... | 25 |
sw_dates_fix-l2 | software engineering | 2 | 0 | A user reported a bug in the `dates` module in the current directory (see ISSUE.md); the existing tests in tests/ still pass. Fix dates.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fix the... | {
"dates.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fromisoform... | {
"dates.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fromisoform... | 25 |
sw_dates_fix-l2 | software engineering | 2 | 1 | A user reported a bug in the `calendar_utils` module in the current directory (see ISSUE.md); the existing tests in tests/ still pass. Fix calendar_utils.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every fu... | {
"calendar_utils.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fr... | {
"calendar_utils.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fr... | 25 |
sw_dates_fix-l2 | software engineering | 2 | 2 | A user reported a bug in the `calendar_utils` module in the current directory (see ISSUE.md); the existing tests in tests/ still pass. Fix calendar_utils.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every fu... | {
"calendar_utils.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fr... | {
"calendar_utils.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fr... | 25 |
sw_dates_fix-l3 | software engineering | 3 | 0 | The `billing_dates` module in the current directory has two bugs: `pytest tests/` fails, and ISSUE.md reports another problem. Fix both in billing_dates.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every fun... | {
"billing_dates.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fro... | {
"billing_dates.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fro... | 25 |
sw_dates_fix-l3 | software engineering | 3 | 1 | The `billing_dates` module in the current directory has two bugs: `pytest tests/` fails, and ISSUE.md reports another problem. Fix both in billing_dates.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every fun... | {
"billing_dates.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fro... | {
"billing_dates.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fro... | 25 |
sw_dates_fix-l3 | software engineering | 3 | 2 | The `calendar_utils` module in the current directory has two bugs: `pytest tests/` fails, and ISSUE.md reports another problem. Fix both in calendar_utils.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every f... | {
"calendar_utils.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fr... | {
"calendar_utils.py": "\"\"\"Date helpers for invoicing. Dates are ISO strings (YYYY-MM-DD).\"\"\"\n\nimport calendar\nfrom datetime import date, timedelta\n\n\ndef add_months(day, n):\n \"\"\"Same day-of-month n months later (n may be negative), clamped to the last day of the target month.\"\"\"\n d = date.fr... | 25 |
sw_report_fix-l1 | software engineering | 1 | 0 | report.py in the current directory produces out/summary.csv from transactions.csv, but the numbers are wrong. Fix report.py so it follows README.md, then run it to regenerate out/summary.csv. Keep `build(in_path, out_path)`; hidden tests also call it on other inputs. | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | 3 |
sw_report_fix-l1 | software engineering | 1 | 1 | report.py in the current directory produces out/summary.csv from transactions.csv, but the numbers are wrong. Fix report.py so it follows README.md, then run it to regenerate out/summary.csv. Keep `build(in_path, out_path)`; hidden tests also call it on other inputs. | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | 3 |
sw_report_fix-l1 | software engineering | 1 | 2 | report.py in the current directory produces out/summary.csv from transactions.csv, but the numbers are wrong. Fix report.py so it follows README.md, then run it to regenerate out/summary.csv. Keep `build(in_path, out_path)`; hidden tests also call it on other inputs. | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | 3 |
sw_report_fix-l2 | software engineering | 2 | 0 | report.py in the current directory produces out/summary.csv from transactions.csv, but the numbers are wrong. Fix report.py so it follows README.md, then run it to regenerate out/summary.csv. Keep `build(in_path, out_path)`; hidden tests also call it on other inputs. | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | 3 |
sw_report_fix-l2 | software engineering | 2 | 1 | report.py in the current directory produces out/summary.csv from transactions.csv, but the numbers are wrong. Fix report.py so it follows README.md, then run it to regenerate out/summary.csv. Keep `build(in_path, out_path)`; hidden tests also call it on other inputs. | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | 3 |
sw_report_fix-l2 | software engineering | 2 | 2 | report.py in the current directory produces out/summary.csv from transactions.csv, but the numbers are wrong. Fix report.py so it follows README.md, then run it to regenerate out/summary.csv. Keep `build(in_path, out_path)`; hidden tests also call it on other inputs. | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | 3 |
sw_report_fix-l3 | software engineering | 3 | 0 | report.py in the current directory produces out/summary.csv from transactions.csv, but the numbers are wrong. Fix report.py so it follows README.md and NOTICE.md, then run it to regenerate out/summary.csv. Keep `build(in_path, out_path)`; hidden tests also call it on other inputs. | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n if row.get(\"statu... | 3 |
sw_report_fix-l3 | software engineering | 3 | 1 | report.py in the current directory produces out/summary.csv from transactions.csv, but the numbers are wrong. Fix report.py so it follows README.md and NOTICE.md, then run it to regenerate out/summary.csv. Keep `build(in_path, out_path)`; hidden tests also call it on other inputs. | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n if row.get(\"statu... | 3 |
sw_report_fix-l3 | software engineering | 3 | 2 | report.py in the current directory produces out/summary.csv from transactions.csv, but the numbers are wrong. Fix report.py so it follows README.md and NOTICE.md, then run it to regenerate out/summary.csv. Keep `build(in_path, out_path)`; hidden tests also call it on other inputs. | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n month = row[\"date... | {
"report.py": "\"\"\"Monthly net sales by category. See README.md for the expected output.\"\"\"\n\nimport csv\nimport os\nfrom decimal import Decimal\n\n\ndef build(in_path, out_path):\n groups = {}\n with open(in_path, newline=\"\") as f:\n for row in csv.DictReader(f):\n if row.get(\"statu... | 3 |
sw_logstats_impl-l1 | software engineering | 1 | 0 | Implement `summarize(path)` in access_summary.py in the current directory. The docstring at the top of access_summary.py is the specification; sample.log is an example input and tests/ has one example test. Keep the function name and signature. Hidden tests run summarize on other logs. | {
"access_summary.py": "\"\"\"Summarize request logs.\n\nEach line of a log looks like\n\n 2026-10-03T14:22:05Z level=INFO path=/api/orders status=200 ms=123\n\ni.e. an ISO timestamp ending in Z, then space-separated key=value fields. Blank lines are\nignored. Every non-blank line is a valid log line.\n\nsummarize... | {
"access_summary.py": "\"\"\"Summarize request logs.\n\nEach line of a log looks like\n\n 2026-10-03T14:22:05Z level=INFO path=/api/orders status=200 ms=123\n\ni.e. an ISO timestamp ending in Z, then space-separated key=value fields. Blank lines are\nignored. Every non-blank line is a valid log line.\n\nsummarize... | 4 |
sw_logstats_impl-l1 | software engineering | 1 | 1 | Implement `summarize(path)` in access_summary.py in the current directory. The docstring at the top of access_summary.py is the specification; sample.log is an example input and tests/ has one example test. Keep the function name and signature. Hidden tests run summarize on other logs. | {
"access_summary.py": "\"\"\"Summarize request logs.\n\nEach line of a log looks like\n\n 2026-10-03T14:22:05Z level=INFO path=/api/orders status=200 ms=123\n\ni.e. an ISO timestamp ending in Z, then space-separated key=value fields. Blank lines are\nignored. Every non-blank line is a valid log line.\n\nsummarize... | {
"access_summary.py": "\"\"\"Summarize request logs.\n\nEach line of a log looks like\n\n 2026-10-03T14:22:05Z level=INFO path=/api/orders status=200 ms=123\n\ni.e. an ISO timestamp ending in Z, then space-separated key=value fields. Blank lines are\nignored. Every non-blank line is a valid log line.\n\nsummarize... | 4 |
sw_logstats_impl-l1 | software engineering | 1 | 2 | Implement `summarize(path)` in logstats.py in the current directory. The docstring at the top of logstats.py is the specification; sample.log is an example input and tests/ has one example test. Keep the function name and signature. Hidden tests run summarize on other logs. | {
"logstats.py": "\"\"\"Summarize request logs.\n\nEach line of a log looks like\n\n 2026-10-03T14:22:05Z level=INFO path=/api/orders status=200 ms=123\n\ni.e. an ISO timestamp ending in Z, then space-separated key=value fields. Blank lines are\nignored. Every non-blank line is a valid log line.\n\nsummarize(path)... | {
"logstats.py": "\"\"\"Summarize request logs.\n\nEach line of a log looks like\n\n 2026-10-03T14:22:05Z level=INFO path=/api/orders status=200 ms=123\n\ni.e. an ISO timestamp ending in Z, then space-separated key=value fields. Blank lines are\nignored. Every non-blank line is a valid log line.\n\nsummarize(path)... | 4 |
sw_money_fix-l2 | software engineering | 2 | 0 | A user reported a bug in the `cents` module in the current directory (see ISSUE.md); the existing tests in tests/ still pass. Fix cents.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fix the... | {
"cents.py": "\"\"\"Money helpers for the billing service. Amounts are integer cents.\"\"\"\n\nfrom decimal import Decimal, ROUND_HALF_UP, ROUND_DOWN\n\n\ndef parse_amount(text):\n \"\"\"Parse '$1,234.50', '1234.5', '-$3' or '(12.00)' (parentheses mean negative) into integer cents.\n More than two decimals rou... | {
"cents.py": "\"\"\"Money helpers for the billing service. Amounts are integer cents.\"\"\"\n\nfrom decimal import Decimal, ROUND_HALF_UP, ROUND_DOWN\n\n\ndef parse_amount(text):\n \"\"\"Parse '$1,234.50', '1234.5', '-$3' or '(12.00)' (parentheses mean negative) into integer cents.\n More than two decimals rou... | 28 |
sw_money_fix-l2 | software engineering | 2 | 1 | A user reported a bug in the `cents` module in the current directory (see ISSUE.md); the existing tests in tests/ still pass. Fix cents.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fix the... | {
"cents.py": "\"\"\"Money helpers for the billing service. Amounts are integer cents.\"\"\"\n\nfrom decimal import Decimal, ROUND_HALF_UP, ROUND_DOWN\n\n\ndef parse_amount(text):\n \"\"\"Parse '$1,234.50', '1234.5', '-$3' or '(12.00)' (parentheses mean negative) into integer cents.\n More than two decimals rou... | {
"cents.py": "\"\"\"Money helpers for the billing service. Amounts are integer cents.\"\"\"\n\nfrom decimal import Decimal, ROUND_HALF_UP, ROUND_DOWN\n\n\ndef parse_amount(text):\n \"\"\"Parse '$1,234.50', '1234.5', '-$3' or '(12.00)' (parentheses mean negative) into integer cents.\n More than two decimals rou... | 28 |
sw_money_fix-l2 | software engineering | 2 | 2 | A user reported a bug in the `cents` module in the current directory (see ISSUE.md); the existing tests in tests/ still pass. Fix cents.py. Every function must behave exactly as its docstring describes. Keep the function names and signatures, and do not edit the tests. Hidden tests will check every function, so fix the... | {
"cents.py": "\"\"\"Money helpers for the billing service. Amounts are integer cents.\"\"\"\n\nfrom decimal import Decimal, ROUND_HALF_UP, ROUND_DOWN\n\n\ndef parse_amount(text):\n \"\"\"Parse '$1,234.50', '1234.5', '-$3' or '(12.00)' (parentheses mean negative) into integer cents.\n More than two decimals rou... | {
"cents.py": "\"\"\"Money helpers for the billing service. Amounts are integer cents.\"\"\"\n\nfrom decimal import Decimal, ROUND_HALF_UP, ROUND_DOWN\n\n\ndef parse_amount(text):\n \"\"\"Parse '$1,234.50', '1234.5', '-$3' or '(12.00)' (parentheses mean negative) into integer cents.\n More than two decimals rou... | 28 |
bank_recon-l3 | finance and economics | 3 | 0 | Reconcile August cash. Match bank and book items by ref. Book items missing from the bank are outstanding checks (negative) or deposits in transit (positive). Bank items missing from the book must be booked (book adjustments). adjusted_balance = bank_statement_ending + deposits in transit - outstanding checks (+ any ba... | {
"bank_statement_aug.csv": "date,ref,description,amount\n2025-08-09,DEP419,DEPOSIT,1984.42\n2025-08-22,DEP335,DEPOSIT,14022.33\n2025-08-22,DEP860,DEPOSIT,9018.69\n2025-08-26,CHK2278,CHECK 2278,-2119.59\n2025-08-31,BNK541,INTEREST,24.29\n2025-08-31,BNK742,NSF RETURN,-2243.14\n2025-08-31,BNK970,SVC FEE,-46.72\n2025-09... | {
"answer.json": "{\"adjusted_balance\": 105520.15, \"outstanding_checks\": [], \"deposits_in_transit\": 10657.81, \"book_adjustments\": [\"BNK541\", \"BNK742\", \"BNK970\"]}"
} | 0 |
bank_recon-l3 | finance and economics | 3 | 1 | Reconcile August cash. Match bank and book items by ref. Book items missing from the bank are outstanding checks (negative) or deposits in transit (positive). Bank items missing from the book must be booked (book adjustments). adjusted_balance = bank_statement_ending + deposits in transit - outstanding checks (+ any ba... | {
"bank_statement_aug.csv": "date,ref,description,amount\n2025-08-07,CHK2699,CHECK 2699,-152.08\n2025-08-10,DEP540,DEPOSIT,9752.28\n2025-08-10,DEP864,DEPOSIT,11329.77\n2025-08-11,CHK2701,CHECK 2701,-4507.19\n2025-08-12,CHK2700,CHECK 2700,-913.29\n2025-08-28,DEP112,DEPOSIT,4023.4\n2025-08-29,DEP611,DEPOSIT,8503.07\n20... | {
"answer.json": "{\"adjusted_balance\": 111371.21, \"outstanding_checks\": [\"CHK2698\"], \"deposits_in_transit\": 7299.51, \"book_adjustments\": [\"BNK130\", \"BNK191\", \"BNK362\"]}"
} | 0 |
bank_recon-l3 | finance and economics | 3 | 2 | Reconcile August cash. Match bank and book items by ref. Book items missing from the bank are outstanding checks (negative) or deposits in transit (positive). Bank items missing from the book must be booked (book adjustments). adjusted_balance = bank_statement_ending + deposits in transit - outstanding checks (+ any ba... | {
"bank_statement_aug.csv": "date,ref,description,amount\n2025-08-06,DEP547,DEPOSIT,10650.29\n2025-08-06,DEP705,DEPOSIT,4374.58\n2025-08-11,CHK3833,CHECK 3833,-3664.56\n2025-08-15,DEP450,DEPOSIT,9598.03\n2025-08-25,DEP998,DEPOSIT,14152.76\n2025-08-27,DEP108,DEPOSIT,5719.13\n2025-08-29,CHK3831,CHECK 3831,-7167.54\n202... | {
"answer.json": "{\"adjusted_balance\": 65276.14, \"outstanding_checks\": [\"CHK3834\"], \"deposits_in_transit\": 0.0, \"book_adjustments\": [\"BNK221\", \"BNK280\", \"BNK872\"]}"
} | 0 |
bond_price-l3 | finance and economics | 3 | 0 | Price the bond holding in bond.json. Coupon dates step back from maturity by 12/coupons_per_year months (same day of month). Let prev and next be the coupon dates with prev <= settlement < next. w = actual days(settlement, next) / actual days(prev, next). Per 100 face: c = coupon_pct/coupons_per_year, y = yield_pct/100... | {
"bond.json": "{\n \"isin\": \"XS4463186129\",\n \"coupon_pct\": 7.0,\n \"coupons_per_year\": 1,\n \"maturity\": \"2028-07-15\",\n \"settlement\": \"2025-09-29\",\n \"yield_pct\": 8.707,\n \"face_holding\": 25000\n}\n",
"notice_trade_correction.txt": "Trade correction: the settlement date in bond.json is wrong. Th... | {
"answer.json": "{\"dirty_price\": 97.0087, \"clean_price\": 95.532, \"accrued_interest\": 369.18}"
} | 0 |
bond_price-l3 | finance and economics | 3 | 1 | Price the bond holding in bond.json. Coupon dates step back from maturity by 12/coupons_per_year months (same day of month). Let prev and next be the coupon dates with prev <= settlement < next. w = actual days(settlement, next) / actual days(prev, next). Per 100 face: c = coupon_pct/coupons_per_year, y = yield_pct/100... | {
"bond.json": "{\n \"isin\": \"XS4167896121\",\n \"coupon_pct\": 3.75,\n \"coupons_per_year\": 2,\n \"maturity\": \"2035-12-15\",\n \"settlement\": \"2025-10-25\",\n \"yield_pct\": 4.031,\n \"face_holding\": 10000\n}\n",
"notice_trade_correction.txt": "Trade correction: the settlement date in bond.json is wrong. T... | {
"answer.json": "{\"dirty_price\": 96.6552, \"clean_price\": 95.272, \"accrued_interest\": 138.32}"
} | 0 |
bond_price-l3 | finance and economics | 3 | 2 | Price the bond holding in bond.json. Coupon dates step back from maturity by 12/coupons_per_year months (same day of month). Let prev and next be the coupon dates with prev <= settlement < next. w = actual days(settlement, next) / actual days(prev, next). Per 100 face: c = coupon_pct/coupons_per_year, y = yield_pct/100... | {
"bond.json": "{\n \"isin\": \"XS6331512191\",\n \"coupon_pct\": 3.0,\n \"coupons_per_year\": 1,\n \"maturity\": \"2028-08-15\",\n \"settlement\": \"2025-08-13\",\n \"yield_pct\": 3.198,\n \"face_holding\": 25000\n}\n",
"notice_trade_correction.txt": "Trade correction: the settlement date in bond.json is wrong. Th... | {
"answer.json": "{\"dirty_price\": 102.9973, \"clean_price\": 100.0055, \"accrued_interest\": 747.95}"
} | 0 |
cap_table-l3 | finance and economics | 3 | 0 | Close the Series A in term_sheet.json. Pre-money fully diluted shares S = all shares in cap_table.csv (common plus all options, granted and unallocated), excluding SAFE conversion shares. price_per_share = pre_money_valuation / S, rounded half-up to 4 dp. new_investor_shares = floor(investment / price_per_share). Each ... | {
"cap_table.csv": "holder,class,shares\nFounder A,common,4221000\nFounder B,common,1942000\nFounder C,common,4014000\nEmployees (granted),options,999000\nUnallocated pool,options,656000\nAngel,common,364000\n",
"term_sheet.json": "{\n \"round\": \"Series A\",\n \"pre_money_valuation\": 12000000,\n \"investment\": ... | {
"answer.json": "{\"price_per_share\": 0.9839, \"new_investor_shares\": 9147271, \"founders_pct\": 45.86, \"safe_shares\": 846979}"
} | 0 |
cap_table-l3 | finance and economics | 3 | 1 | Close the Series A in term_sheet.json. Pre-money fully diluted shares S = all shares in cap_table.csv (common plus all options, granted and unallocated), excluding SAFE conversion shares. price_per_share = pre_money_valuation / S, rounded half-up to 4 dp. new_investor_shares = floor(investment / price_per_share). Each ... | {
"cap_table.csv": "holder,class,shares\nFounder A,common,2185000\nFounder B,common,3767000\nFounder C,common,2307000\nEmployees (granted),options,858000\nUnallocated pool,options,695000\nAngel,common,94000\n",
"term_sheet.json": "{\n \"round\": \"Series A\",\n \"pre_money_valuation\": 37000000,\n \"investment\": 1... | {
"answer.json": "{\"price_per_share\": 3.7351, \"new_investor_shares\": 4551417, \"founders_pct\": 52.72, \"safe_shares\": 1207425}"
} | 0 |
cap_table-l3 | finance and economics | 3 | 2 | Close the Series A in term_sheet.json. Pre-money fully diluted shares S = all shares in cap_table.csv (common plus all options, granted and unallocated), excluding SAFE conversion shares. price_per_share = pre_money_valuation / S, rounded half-up to 4 dp. new_investor_shares = floor(investment / price_per_share). Each ... | {
"cap_table.csv": "holder,class,shares\nFounder A,common,4520000\nFounder B,common,1997000\nFounder C,common,1702000\nEmployees (granted),options,439000\nUnallocated pool,options,116000\nAngel,common,377000\n",
"term_sheet.json": "{\n \"round\": \"Series A\",\n \"pre_money_valuation\": 12000000,\n \"investment\": ... | {
"answer.json": "{\"price_per_share\": 1.3113, \"new_investor_shares\": 2287805, \"founders_pct\": 71.85, \"safe_shares\": 0}"
} | 0 |
covenant-l3 | finance and economics | 3 | 0 | Test the borrower's covenants per credit_agreement.txt at the test quarter in balance_sheet.json. Compliant means both covenant tests pass on the unrounded ratios. Save the answer as a JSON object in answer.json: {"net_leverage": ratio rounded to 2 dp, "interest_coverage": ratio rounded to 2 dp, "compliant": true/false... | {
"quarterly_pnl.csv": "quarter,revenue,cogs,opex_incl_restructuring,d_and_a,interest_expense,restructuring_charges,tax_expense\n2024Q1,15700000,8937035,2799494,604023,935811,0,208228\n2024Q2,14200000,6775889,4147759,716537,891452,442661,530771\n2024Q3,14700000,7231757,3694153,505282,1222731,0,278625\n2024Q4,9800000,... | {
"answer.json": "{\"net_leverage\": 3.15, \"interest_coverage\": 3.16, \"compliant\": false}"
} | 0 |
covenant-l3 | finance and economics | 3 | 1 | Test the borrower's covenants per credit_agreement.txt at the test quarter in balance_sheet.json. Compliant means both covenant tests pass on the unrounded ratios. Save the answer as a JSON object in answer.json: {"net_leverage": ratio rounded to 2 dp, "interest_coverage": ratio rounded to 2 dp, "compliant": true/false... | {
"quarterly_pnl.csv": "quarter,revenue,cogs,opex_incl_restructuring,d_and_a,interest_expense,restructuring_charges,tax_expense\n2024Q1,13600000,7075167,3214438,487133,870308,0,406822\n2024Q2,8700000,4326008,1343979,494440,547190,547036,247638\n2024Q3,17100000,9346998,4195367,1016545,1119005,738870,374482\n2024Q4,820... | {
"answer.json": "{\"net_leverage\": 3.66, \"interest_coverage\": 3.94, \"compliant\": true}"
} | 0 |
covenant-l3 | finance and economics | 3 | 2 | Test the borrower's covenants per credit_agreement.txt at the test quarter in balance_sheet.json. Compliant means both covenant tests pass on the unrounded ratios. Save the answer as a JSON object in answer.json: {"net_leverage": ratio rounded to 2 dp, "interest_coverage": ratio rounded to 2 dp, "compliant": true/false... | {
"quarterly_pnl.csv": "quarter,revenue,cogs,opex_incl_restructuring,d_and_a,interest_expense,restructuring_charges,tax_expense\n2024Q1,7000000,3467689,1301598,259022,571812,0,118558\n2024Q2,8100000,3685416,1893958,294047,643742,0,242658\n2024Q3,8700000,4030958,1574744,496859,584559,0,188225\n2024Q4,15800000,8657819,... | {
"answer.json": "{\"net_leverage\": 2.53, \"interest_coverage\": 6.85, \"compliant\": true}"
} | 0 |
fx_consolidation-l2 | finance and economics | 2 | 0 | Consolidate the group's 2025 trial balances into USD. revenue and expenses translate at the currency's simple mean of the 12 monthly 'average' rates; cash and receivables translate at the 'closing' rate. USD amounts need no translation. net income = revenue - expenses. Eliminate intercompany sales: subtract the transla... | {
"entities.csv": "entity,functional_currency\nUS-PARENT,USD\nSUB-GBP,GBP\nSUB-INR,INR\nSUB-CAD,CAD\nSUB-BRL,BRL\n",
"trial_balance_2025.csv": "entity,account,amount_local\nUS-PARENT,revenue,8065563.44\nUS-PARENT,expenses,1768323.6\nUS-PARENT,cash,7622481.17\nUS-PARENT,receivables,6904180.54\nSUB-GBP,revenue,567206... | {
"answer.json": "{\"revenue_usd\": 28399844.94, \"net_income_usd\": 17231931.2, \"cash_usd\": 21517487.84}"
} | 0 |
fx_consolidation-l2 | finance and economics | 2 | 1 | Consolidate the group's 2025 trial balances into USD. revenue and expenses translate at the currency's simple mean of the 12 monthly 'average' rates; cash and receivables translate at the 'closing' rate. USD amounts need no translation. net income = revenue - expenses. Eliminate intercompany sales: subtract the transla... | {
"entities.csv": "entity,functional_currency\nUS-PARENT,USD\nSUB-BRL,BRL\nSUB-CHF,CHF\nSUB-AUD,AUD\nSUB-JPY,JPY\n",
"trial_balance_2025.csv": "entity,account,amount_local\nUS-PARENT,revenue,6421180.01\nUS-PARENT,expenses,2694041.11\nUS-PARENT,cash,7420921.93\nUS-PARENT,receivables,5548546.91\nSUB-BRL,revenue,15645... | {
"answer.json": "{\"revenue_usd\": 27148701.45, \"net_income_usd\": 11755174.5, \"cash_usd\": 17170226.24}"
} | 0 |
fx_consolidation-l2 | finance and economics | 2 | 2 | Consolidate the group's 2025 trial balances into USD. revenue and expenses translate at the currency's simple mean of the 12 monthly 'average' rates; cash and receivables translate at the 'closing' rate. USD amounts need no translation. net income = revenue - expenses. Eliminate intercompany sales: subtract the transla... | {
"entities.csv": "entity,functional_currency\nUS-PARENT,USD\nSUB-BRL,BRL\nSUB-GBP,GBP\nSUB-EUR,EUR\nSUB-CHF,CHF\n",
"trial_balance_2025.csv": "entity,account,amount_local\nUS-PARENT,revenue,7674573.65\nUS-PARENT,expenses,1910105.96\nUS-PARENT,cash,967247.33\nUS-PARENT,receivables,1433943.9\nSUB-BRL,revenue,2621632... | {
"answer.json": "{\"revenue_usd\": 27423092.81, \"net_income_usd\": 14305986.66, \"cash_usd\": 21712865.27}"
} | 0 |
payroll-l2 | finance and economics | 2 | 0 | Run the 2025-11 payroll (one of 12 monthly periods) from employees.csv and payroll_policy.json. Per employee, every amount rounded half-up to cents: gross = annual_salary/12; 401k = gross*pretax_401k_pct/100; income tax = (progressive tax on max(0, (gross-401k)*12 - standard_deduction_annual), each bracket's rate apply... | {
"employees.csv": "emp_id,annual_salary,pretax_401k_pct,ytd_gross_before_period\nE2120,360000,8,300000.0\nE6571,227000,5,189166.67\nE3362,164000,3,136666.67\nE7643,373000,0,310833.33\nE6164,400000,10,333333.33\n",
"payroll_policy.json": "{\n \"pay_period\": \"2025-11\",\n \"periods_per_year\": 12,\n \"standard_ded... | {
"answer.json": "{\"total_net_pay\": 88461.26, \"total_income_tax\": 27838.25, \"total_social_security\": 847.33}"
} | 0 |
payroll-l2 | finance and economics | 2 | 1 | Run the 2025-11 payroll (one of 12 monthly periods) from employees.csv and payroll_policy.json. Per employee, every amount rounded half-up to cents: gross = annual_salary/12; 401k = gross*pretax_401k_pct/100; income tax = (progressive tax on max(0, (gross-401k)*12 - standard_deduction_annual), each bracket's rate apply... | {
"employees.csv": "emp_id,annual_salary,pretax_401k_pct,ytd_gross_before_period\nE4030,396000,10,330000.0\nE4131,61000,5,50833.33\nE7412,230000,10,191666.67\nE9603,46000,5,38333.33\nE6084,385000,8,320833.33\nE3365,214000,8,178333.33\n",
"payroll_policy.json": "{\n \"pay_period\": \"2025-11\",\n \"periods_per_year\... | {
"answer.json": "{\"total_net_pay\": 76761.0, \"total_income_tax\": 21737.55, \"total_social_security\": 552.84}"
} | 0 |
payroll-l2 | finance and economics | 2 | 2 | Run the 2025-11 payroll (one of 12 monthly periods) from employees.csv and payroll_policy.json. Per employee, every amount rounded half-up to cents: gross = annual_salary/12; 401k = gross*pretax_401k_pct/100; income tax = (progressive tax on max(0, (gross-401k)*12 - standard_deduction_annual), each bracket's rate apply... | {
"employees.csv": "emp_id,annual_salary,pretax_401k_pct,ytd_gross_before_period\nE5850,381000,10,317500.0\nE7481,321000,6,267500.0\nE1282,149000,10,124166.67\n",
"payroll_policy.json": "{\n \"pay_period\": \"2025-11\",\n \"periods_per_year\": 12,\n \"standard_deduction_annual\": 15000,\n \"federal_brackets_annual\... | {
"answer.json": "{\"total_net_pay\": 47932.91, \"total_income_tax\": 14637.46, \"total_social_security\": 769.83}"
} | 0 |
pe_waterfall-l3 | finance and economics | 3 | 0 | Split exit_proceeds between LPs and the GP. Tier 1: LPs get back total called capital. Tier 2: LPs get the preferred return: for each call, amount * ((1 + pref/100)^(days/365) - 1), days = actual days from call_date to exit_date. Tier 3 (full catch-up): the GP gets 100% until it holds carried_interest_pct of tiers 2+3,... | {
"lpa_terms.json": "{\n \"preferred_return_pct\": 8,\n \"carried_interest_pct\": 20,\n \"gp_catch_up\": \"full\",\n \"exit_date\": \"2025-06-30\",\n \"exit_proceeds\": 28408293.73\n}\n",
"capital_calls.csv": "call_date,amount\n2019-08-01,2800000\n2020-11-19,2300000\n2021-03-20,4800000\n",
"notice_side_letter.txt... | {
"answer.json": "{\"lp_total\": 24796634.98, \"gp_carry\": 3611658.75}"
} | 0 |
pe_waterfall-l3 | finance and economics | 3 | 1 | Split exit_proceeds between LPs and the GP. Tier 1: LPs get back total called capital. Tier 2: LPs get the preferred return: for each call, amount * ((1 + pref/100)^(days/365) - 1), days = actual days from call_date to exit_date. Tier 3 (full catch-up): the GP gets 100% until it holds carried_interest_pct of tiers 2+3,... | {
"lpa_terms.json": "{\n \"preferred_return_pct\": 8,\n \"carried_interest_pct\": 25,\n \"gp_catch_up\": \"full\",\n \"exit_date\": \"2025-06-30\",\n \"exit_proceeds\": 20343502.76\n}\n",
"capital_calls.csv": "call_date,amount\n2021-02-01,4400000\n2022-06-08,2900000\n",
"notice_side_letter.txt": "Administrator co... | {
"answer.json": "{\"lp_total\": 17095127.07, \"gp_carry\": 3248375.69}"
} | 0 |
End of preview. Expand in Data Studio
Finance Desk: OpenEnv environment tasks
Public task preview and source for the Finance Desk OpenEnv Arena environment.
v3: terminal mode (current)
- Image (public, linux/amd64):
ghcr.io/bp-high/finance-desk:v3 - Source: https://github.com/bp-high/openenv_financial_task_env/tree/dev/pensive-lovelace-bt5rsn/finance_desk
- Contract: OpenEnv's
tbench2_env(Terminal-Bench 2). The action and observation models are copied field for field, and/schemamatches the operator's TB2 smoke test exactly. Actions:exec(run a shell command in the workdir),write_file,evaluate(grade and end the episode). Every observation carries the instruction. - Budgets per task: 4096 completion / 8192 context tokens, 32 tool calls. These are the same as the TB2 smoke test.
Each episode builds a fresh workdir (/app) holding a small repo or a set of data files, and gives a prose instruction.
Software (12 task ids):
- fix seeded bugs in a stats, money or dates module (a failing visible test, a bug report, or both);
- implement a log summarizer to a docstring spec;
- repair a report script and regenerate its output.
Hidden tests run against the agent's code. At generation time the shipped code is proven to fail them and the reference to pass.
Finance (30 task ids): procedural finance problems across 13 families (AR aging, bank reconciliation, bond pricing, cap table, covenants, DCF/NPV, depreciation, FX consolidation, loan amortization, payroll, PE waterfall, portfolio return, real growth). The answer goes into
answer.jsonand is graded field by field.
Reward is all-or-nothing: 1.0 if the hidden tests or every answer field pass, else 0.0.
Every task id was calibrated on live Qwen3.8-27B with an arena-style loop. Only mixed-outcome ids are submitted (v3b: 42 of 57; v3: 19). The
per-id table is in calibration_v3.md, and why terminal mode is explained in STRATEGY_v3.md (both in this dataset).
tasks.jsonl has three illustrative instances per submitted task id: instruction, the workdir files, the
reference solution oracle, domain, level, seed. Hidden test cases are regenerated per instance at train
time. The container makes fresh instances during training, so this preview is illustrative, not the trained set.
v2 (previous)
v2/tasks.jsonl is the v2 preview: single-step, with the files inlined in the first observation, across all eight
domains. Image ghcr.io/bp-high/finance-desk:v2. The ABC-notation (music) family in v2 is derived from
XiaomiMiMo/MiMo-V2.6-RL-oss (Apache-2.0).
source/fdesk/ is the exact environment package (v2 under fdesk/, v3 under fdesk/terminal/).
- Downloads last month
- 156