compare_studies
ChatGPTCompare two published GreyScienx studies. Returns each question, shared topics and geographies, assumption names used by both, and any published contradiction or qualification between them. The comparison does not create a new finding.
find_counterarguments
ChatGPTFind published contradictions, qualifications, and the study's own limitations for a GreyScienx study or one of its claims. External critiques are included only when a person has published one.
get_assumptions
ChatGPTGet the assumption objects for a GreyScienx study. Use when an answer depends on what the model assumed. Until extraction, this returns a pending state rather than invented parameters.
get_claim
ChatGPTGet one claim from a GreyScienx study, including evidence status, dependencies, and limitations when that claim has been extracted. Until extraction, this returns an explicit pending state and no conclusion.
get_manuscript
ChatGPTGet the complete Markdown manuscript for a GreyScienx study, or one named section. This is the full source text behind the PDF, not a generated summary. Omit section for the whole manuscript; use a heading title or heading slug for a section.
get_model_spec
ChatGPTGet the registered model statements for a GreyScienx study. Statements come from the manuscript. An empty list means no model object is registered.
get_profile
ChatGPTGet a GreyScienx author profile: name, affiliation, full biography, statements, absences, and the published studies attributed to that person. Omit profile_id to list profiles. Use when a citation needs the author record. Do not fill fields the profile marks as not published.
get_related_research
ChatGPTGet other GreyScienx studies in the same series. Use when a question sits next to a study the user has already found.
get_reproducibility_package
ChatGPTGet the reproducibility record for a GreyScienx study: whether code and data are recorded, the published files, and the reproduce page. A null repository URL means no separate public code repository is published.
get_research_graph
ChatGPTGet the GreyScienx claim graph: claims depend on assumptions, cite modelled results, and are generated by a model. Pass a study id to limit the graph to one study. Omit it for the whole corpus. Edges are relationships, not extra findings.
get_results
ChatGPTGet modelled results for a GreyScienx study, with evidence status. Until extraction, this returns a pending state and does not paraphrase the summary as a result.
get_series
ChatGPTGet a GreyScienx series, its question, and the studies in it. Omit series_slug to list every series.
get_sources
ChatGPTGet the manuscript and any extracted sources for a GreyScienx study. The manuscript PDF is available now.
get_study
ChatGPTGet one GreyScienx study by stable id (for example GSX:POP-009) or paper slug. Returns the question, catalog summary, epistemic status, manuscript link, and related studies. The summary is the catalog description, not a conditional finding.
get_topic
ChatGPTGet a GreyScienx topic hub: definition, connected topics, and the studies filed under it. Omit topic_slug to list topics.
render_analysis_summary
ChatGPTRender a compact GreyScienx analysis-summary widget for one published study. First call get_study and, when a claim is central to the answer, get_claim or trace_claim; then pass the selected study id and optional claim id here. The widget shows the finding, evidence status, sensitivity, evidence counts, main limitation, freshness and a handoff to the full study.
search_research
ChatGPTSearch GreyScienx exploratory research, mathematical models, simulations, counterfactual studies, socioeconomic analyses, demographic models, and public-policy scenarios. Matches questions, claims, assumptions, results, and limitations, including related wording such as houses for housing. Use when the user asks whether a question has been modelled or wants the assumptions behind a modelled claim. Returns the study, its evidence status, the matching claim, and the key assumptions. It does not return a naked conclusion.
trace_claim
ChatGPTTrace one GreyScienx claim to the assumptions it depends on, the modelled results it cites, the model that generated it, and its main limitation. Use this instead of quoting a conclusion alone.
validate_critique
ChatGPTCheck a critique of a published GreyScienx study or claim. The critique names its target, type, argument, evidence, and proposed change. The tool does not store or publish the critique.
validate_fork
ChatGPTCheck a fork of a published GreyScienx study. A fork records the parent study, what changed, the new results, and a conclusion with its evidence status and limitation. The tool does not store or publish the fork.
validate_revision
ChatGPTCheck a proposed revision of a published GreyScienx study. It names the study, the version it starts from, and the change. The tool does not store or publish the revision.
validate_submission
ChatGPTCheck a proposed GreyScienx research object. Requires a question, research type, evidence status, and an explicit AI-assistance disclosure. The tool does not store or publish the object. A valid result still waits for human review.