Data catalogue

74 results

• retrieval status

  Dataset Builder

  Per-task status, calls used, values accepted/rejected, and unavailable sources.

  capability

• run research

  Insight Generation

  Run research briefs as bounded sub-agents (default: all planned briefs, in parallel). Each uses only its allowed sources; every evidence item passes the scope gate.

  capability

• run retrieval

  Dataset Builder

  Run retrieval tasks as bounded agents (default: all planned tasks, in parallel). Every recorded value passes the value gate; rejected values are parked with a reason.

  capability

• select findings

  Insight Generation

  Keep only findings that serve a Charter sub-question (the rest are parked for the appendix). `keep`/`park` override the decision for specific finding ids afterwards.

  capability

• set value

  Dataset Builder

  Manual override for one cell (value=None clears the override). Logged with the note and shown as 'override' in the workbook.

  capability

• start dataset job

  Dataset Builder

  Start a full build in the background and return at once. Pass `dataset_id` to build a dataset whose spec you already reviewed with define_dataset_spec. Poll get_dataset_job(dataset_id) — it long-polls.

  capability

• start insight article

  Insight Generation

  Start generate_insight_article as a background job and return at once with its session_id. Poll get_article_job(session_id) every ~20s for stage and progress; the finished article is in its `result`. Same inputs as generate_insight_article.

  capability

• submit research result

  Insight Generation

  Submit a host sub-agent's evidence for a brief. Each item is scope-gated; rejected items are parked with the reason. final=True (default) also closes the brief; its verdicts/answers merge with any earlier submit (newest wins per hypothesis / sub-question). final=False records the evidence only — use it to submit as you go and read each gate verdict, then close the brief with one final submit (evid

  capability

• submit values

  Dataset Builder

  Host mode: submit values for a task — each {entity, metric, period, value, unit, basis, source_type, source_url|source_ref, source_title, published, locator}, AS PRINTED. Each is gated; rejected ones come back with the reason.

  capability

• test hypothesis

  Insight Generation

  Test one hypothesis: timing, magnitude, signatures in the user's data, source independence, contradictions → grade and the causal wording it allows.

  capability

• to insight article

  Dataset Builder

  Hand the finished dataset to the Insight Generation MCP to research and write an article answering `question`. Returns its job's session_id (poll it there with get_article_job).

  capability

• update dataset spec

  Dataset Builder

  Edit the Spec: top-level keys in `patch` replace the stored ones (lists replace wholesale), e.g. {"entities": [...]}, {"periods": {"start": "Q1 FY22"}}, {"metrics": [...]}, {"conflict_tolerance_pct": 1}. Re-validated on save.

  capability

• validate dataset

  Dataset Builder

  Completeness per metric, derived-vs-reported mismatches, range checks, fraction-vs- percent slips, scale jumps and outliers against each series' own history.

  capability

• verify article

  Insight Generation

  Check the draft: every number traces to a fact or evidence, citations exist, causal wording ≤ grade, core question answered up front, every sub-question covered, nothing off-scope. revise=True applies one fix-up pass and re-checks.

  capability

Inventory snapshot 15 Nov 2026. Publisher dates and individual source checks can differ.