Cited research reports
Write a sourced report whose claims are each backed by a cited page.
20 project-authored proposals · taxonomy v1.1 · 0 new benchmark runs. Verifiers and partial-credit rules below are designs, not executed results.
- Typical applications
- Chrome · Word · Google Docs
- Class GUI-need label
AP· API-sufficient
Digital-employee context
Project class assignments, not demonstrated employee capability or imported benchmark coverage. Acceptance criteria and value metrics are proposals; no ROI has been measured.
Related roles
- Projects, research and collaboration · Project records, dependency schedules, cited reports, confirmed colleague input and conflict-free shared work are distinct outcomes.
Task proposals · 20
Expand to inspect the complete original wording and proposed checks. Original proposal text remains English in both interfaces.
R04.01.01.03-T01 Compare payer prior-authorization policies
Original instruction
Write report.docx comparing the published prior-authorization requirements for a named drug class across three large US health plans. Cover: covered indications, step-therapy requirements, and appeal process, each claim backed by a citation to the plan's own policy page. >=10 distinct citations. 800-1200 words. This is a policy comparison, not medical advice.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 4 / 5
Proposed verifier
Every flagged claim has a resolvable citation; >=10 distinct sources across the 3 plans; required sections present; word count in range. Citation support is checked by a FACT-style rubric judge (stated here because no deterministic support check exists).
Proposed partial credit
30% all 3 plans covered; 30% required subsections; 30% >=10 resolvable citations; 10% length
R04.01.01.03-T02 Battery price-trend brief
Original instruction
Write brief.docx on lithium-iron-phosphate battery pack price trends over the last five years: include at least four sourced numeric data points (year + $/kWh), the main drivers, and an outlook. Each figure cited to its source. 600-900 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
At least 4 numeric data points each with a resolvable citation; drivers and outlook sections present; figures consistent with sources (rubric judge checks figure-to-source support). Length in range.
Proposed partial credit
40% 4+ cited data points; 30% drivers+outlook sections; 20% citation accuracy; 10% length
R04.01.01.03-T03 Compare three national AI-safety bills
Original instruction
Write report.docx comparing three named national/regional AI-governance proposals on: scope, risk tiers, obligations for providers, and enforcement. Every substantive claim must carry a section-level citation to the bill text or an official summary. >=15 citations. 1500-2000 words with a comparison table.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word · Google Docs
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
long· Long-horizon- Proposed difficulty
- 5 / 5
Proposed verifier
All four dimensions covered for all three bills in a table; >=15 resolvable section-level citations; claim-citation coverage >= 0.9 (rubric judge verifies support against cited sections). Length in range.
Proposed partial credit
30% all bills x dimensions; 25% comparison table; 35% citation coverage+accuracy; 10% length
R04.01.01.03-T04 Mini literature review
Original instruction
Write litreview.docx summarizing the state of research on a specified narrow technical topic: synthesize at least six peer-reviewed or preprint sources, group them by approach, and note open questions. Each source cited with a resolvable link. 900-1300 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
>=6 distinct scholarly sources each cited and resolvable; grouped-by-approach structure present; open-questions section present; a rubric judge checks that summaries match the cited sources. Length in range.
Proposed partial credit
40% 6+ cited sources; 30% grouped synthesis; 20% open questions; 10% length
R04.01.01.03-T05 Compare privacy laws across jurisdictions
Original instruction
Write report.docx comparing the data-protection regimes of three named jurisdictions on: lawful bases, data-subject rights, breach-notification timelines, and penalties. Each statement cited to the statute or regulator guidance. >=12 citations. 1200-1600 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word · Google Docs
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 4 / 5
Proposed verifier
All four dimensions for all three jurisdictions; >=12 resolvable citations; claim-citation coverage >= 0.9 (rubric judge). Length in range.
Proposed partial credit
35% jurisdictions x dimensions; 35% citation coverage+accuracy; 20% structure; 10% length
R04.01.01.03-T06 Industry market overview
Original instruction
Write overview.docx giving a neutral market overview of a specified industry: market size, top segments, growth rate, and two or three key players, with every figure cited to its source. 600-900 words. No investment advice.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Market-size, segments, growth, and players sections present; each figure carries a resolvable citation; rubric judge checks figure-source support. Length in range.
Proposed partial credit
40% required sections; 30% cited figures; 20% citation accuracy; 10% length
R04.01.01.03-T07 Technology-development timeline
Original instruction
Write timeline.docx: a one-page cited timeline of a specified technology's development, with at least eight dated milestones, each with a one-line description and a resolvable citation. 400-600 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 2 / 5
Proposed verifier
>=8 dated milestones in chronological order, each with a resolvable citation; dates consistent with sources (rubric spot-check). Length in range.
Proposed partial credit
50% 8+ dated milestones; 30% citations resolvable; 20% chronological + length
R04.01.01.03-T08 Sourced product comparison
Original instruction
Write compare.docx comparing three named products on five specifications (from each product's official spec page), with a table and a short recommendation for two use cases. Every spec value cited to the manufacturer page. 500-800 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word · Google Docs
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Comparison table with all three products x five specs; each spec value cited and matching the source page (rubric check); recommendation section present. Length in range.
Proposed partial credit
40% complete spec table; 30% cited values accurate; 20% recommendation; 10% length
R04.01.01.03-T09 Policy brief on a public issue
Original instruction
Write brief.docx: a policy brief on a specified public-interest issue with Background, Current Policy, Options, and Recommendation sections. Each factual claim cited; include at least three primary/official sources. 900-1300 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word · Google Docs
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
All four sections present; >=3 primary/official sources among the citations; claim-citation coverage >= 0.85 (rubric judge). Length in range.
Proposed partial credit
30% four sections; 30% primary sources; 30% citation coverage; 10% length
R04.01.01.03-T10 Cited concept explainer
Original instruction
Write explainer.docx: a clear 1-page explanation of a specified scientific or technical concept for a general audience, with at least four claims each backed by a resolvable citation to a reputable source. 350-550 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & forms- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 2 / 5
Proposed verifier
>=4 cited claims each resolvable; a rubric judge checks that the explanation is accurate and the citations support it. Length in range.
Proposed partial credit
50% 4+ cited claims; 30% citation accuracy; 20% clarity + length
R04.01.01.03-T11 Renewables adoption comparison
Original instruction
Write report.docx comparing renewable-energy adoption across three named countries: share of generation, leading technology, and one policy driver each, every figure cited. Include a small table and a 150-word synthesis. 900-1300 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word · Google Docs
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 4 / 5
Proposed verifier
Table of the three countries x metrics with cited figures; policy driver per country cited; synthesis present; rubric judge verifies figure-source support. Length in range.
Proposed partial credit
35% countries x metrics; 35% cited figures accurate; 20% synthesis; 10% length
R04.01.01.03-T12 Summarize three named studies
Original instruction
Write summary.docx: for three named studies/reports, give each a paragraph covering method, key finding, and one limitation, each with a resolvable citation, then a short comparison. 700-1000 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Each study has method/finding/limitation with a resolvable citation; comparison paragraph present; rubric judge checks summaries match the sources. Length in range.
Proposed partial credit
40% all three summarized; 30% citations; 20% comparison; 10% length
R04.01.01.03-T13 Short cited regulation FAQ
Original instruction
Write faq.docx: a 6-question FAQ about a specified public regulation (who it applies to, key obligations, deadlines, penalties), each answer carrying a citation to the official text or regulator guidance. 400-600 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word · Google Docs
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & forms- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 2 / 5
Proposed verifier
Six Q&A pairs, each with a resolvable official citation; a rubric judge checks each answer is supported by its citation. Length in range.
Proposed partial credit
50% six answered; 30% official citations; 20% accuracy + length
R04.01.01.03-T14 Sector trends overview (neutral)
Original instruction
Write trends.docx: a neutral overview of three current trends in a specified sector, each with two cited data points or examples, and a closing paragraph on uncertainties. No recommendations to buy/sell. 700-1000 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Three trends each with two resolvable citations; uncertainties paragraph present; rubric judge checks support and neutrality. Length in range.
Proposed partial credit
40% three trends; 30% 2 cited points each; 20% uncertainties; 10% length
R04.01.01.03-T15 Compare three transit systems
Original instruction
Write report.docx comparing the public-transit systems of three named cities on: network length, ridership, fare, and one accessibility feature, with every figure cited. Include a table. 800-1100 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word · Google Docs
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Table of three cities x four metrics with cited figures; accessibility feature per city cited; rubric judge verifies figures. Length in range.
Proposed partial credit
40% cities x metrics; 35% cited figures accurate; 15% table; 10% length
R04.01.01.03-T16 Historical cause analysis with primary sources
Original instruction
Write essay.docx analyzing the main causes of a specified well-documented historical event, drawing on at least three primary sources and three secondary sources, each claim cited. 1200-1600 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
long· Long-horizon- Proposed difficulty
- 4 / 5
Proposed verifier
>=3 primary and >=3 secondary sources cited and resolvable; claim-citation coverage >= 0.85 (rubric judge); causal argument structured. Length in range.
Proposed partial credit
30% source mix; 30% citation coverage; 30% argument structure; 10% length
R04.01.01.03-T17 Climate adaptation measures report
Original instruction
Write report.docx surveying climate-adaptation measures for a specified hazard in a named region: list at least five measures, each with evidence of effectiveness and a resolvable citation, plus cost considerations. 1000-1400 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word · Google Docs
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
long· Long-horizon- Proposed difficulty
- 4 / 5
Proposed verifier
>=5 measures each with an effectiveness claim + resolvable citation; cost section present; rubric judge checks support. Length in range.
Proposed partial credit
40% 5+ measures cited; 30% effectiveness evidence; 20% cost section; 10% length
R04.01.01.03-T18 Compare three open standards
Original instruction
Write compare.docx comparing three named open technical standards on: governing body, versioning, key features, and adoption, each statement cited to the spec or official site. Include a table. 800-1100 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Table of three standards x four dimensions with resolvable citations; rubric judge checks claim support. Length in range.
Proposed partial credit
40% standards x dimensions; 35% citation coverage+accuracy; 15% table; 10% length
R04.01.01.03-T19 Deep multi-section sourced report
Original instruction
Write report.docx answering a complex specified question with sections: Executive Summary, Background, Analysis (3 sub-questions), Counterarguments, and Conclusion. >=20 distinct resolvable citations, with an endnote reference list. 2500-3500 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word · Google Docs
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
long· Long-horizon- Proposed difficulty
- 5 / 5
Proposed verifier
All required sections and the three sub-questions present; >=20 distinct resolvable citations with a reference list; claim-citation coverage >= 0.9 and a RACE-style rubric scores depth/coverage/clarity. Length in range.
Proposed partial credit
25% section structure; 25% >=20 citations; 30% coverage+accuracy; 20% rubric quality
R04.01.01.03-T20 Balanced evidence analysis of a debate
Original instruction
Write analysis.docx weighing the evidence on a debated public-policy question: present the strongest cited case for each side (>=4 citations per side), identify where the evidence is contested, and give a reasoned, balanced conclusion. 1500-2200 words.
- Platforms
web· Any OS (web)win· Windowsmac· macOS- Applications
- Chrome · Word · Google Docs
- Interfaces
brow· Browserdesk· Desktop GUI- GUI need · task setting
AP· API-sufficient- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
nav· Pan/zoom/scrolltype· Text & formsmulti· Cross-app- Work product
WP01· Document- Conditions
long· Long-horizonclarify· Ask when unclear- Proposed difficulty
- 5 / 5
Proposed verifier
>=4 resolvable citations per side; a contested-evidence section; a rubric judge scores balance, citation support (>=0.85 coverage), and that the conclusion follows from the evidence. Length in range.
Proposed partial credit
30% both sides with 4+ citations; 25% contested section; 30% balance+support rubric; 15% conclusion+length
Mapped research references
Existing benchmark-to-class links from the research catalog. They are not record-level annotations, executions or product endorsements.