Build a Policy Meeting Information Packet Assembler — Implementation planning and evidence-linked development

## User
A research team preparing a policy committee briefing.

## Build
Prepare a dated macroeconomic briefing for a simulated policy meeting. Track what was available, what was revised, and which statements lack supporting evidence.

## Input Contract
- Key: meeting_id + evidence_id.
- Fields: observation_period, released_at, source_url, claim_id, reviewer_state.
- Separate observation dates, publication dates, and meeting dates.

## Workflow
1. Freeze the permitted evidence set at each meeting cutoff.
2. Calculate a documented set of briefing indicators.
3. Attach sources to draft claims and identify unsupported statements.
4. Export a reviewable packet and immutable evidence manifest.

## Interface
A briefing editor with claim-to-source links, stale-series badges, and explicit review approvals.

## Deliverables
- A reproducible briefing compiler.
- A claim ledger showing supported, disputed, and unsupported statements.
- Runnable local application, setup commands, and a dependency lockfile.
- Versioned input data, data dictionary, acquisition instructions, and rights notes.
- Calculation traces, source references, assumption records, and exported review artifacts.
- A deterministic baseline and tests for normal, incomplete, contradictory, and failure-recovery cases.
- A recorded complete workflow with a reproducible manifest and explicit limitations.

## Acceptance Criteria
- An unpublished release cannot support a meeting claim.
- Every numerical sentence resolves to a calculation and source.
- Conflicting sources remain visible to the reviewer.
- Later annotations cannot alter an approved snapshot.

## Contrast Case
The same statistical value appears in two documents, but only one existed at the meeting time. Only that document is eligible.

## Dataset Scope
The attached starter files are prototype inputs. Expected starter answers are visible. They are not independent holdouts.

Full target: Create twelve historical meeting cutoffs and a release calendar. Pair them with public economic releases and fictional analyst annotations.

Complete target contracts and expert review remain pending. Public observations are not historical release vintages.

## Tools
DuckDB, Pydantic, Jinja2

Use an available model appropriate to the task. Record provider, model identifier, version or retrieval date, settings, cost, and latency. Use deterministic code for calculations. Models may extract, retrieve, or explain evidence. They must not invent missing records, override approval rules, or claim professional authority.

## Constraints
The application prepares evidence. It does not represent a central bank or prescribe policy.

- No fabricated records, source citations, professional endorsements, or expert-adjudication claims.
- No future-data leakage into dated decisions, model fitting, or historical valuation inputs.
- No unauthorized payments, account restrictions, regulatory filings, or production changes.
- No automated appraisal certification, tax-compliance determination, or unsupported fraud accusation.
- No concealed uncertainty or silently invented inputs required for a numerical conclusion.

## Scoring
- Correct domain calculations and justified methods with documented assumptions. Weight: 30.
- Source lineage, temporal integrity, rights checks, and explicit synthetic-data labels. Weight: 25.
- A working application with review, export, error handling, and reproducible state. Weight: 20.
- Adversarial cases, uncertainty, missing-data behavior, and sensitivity analysis. Weight: 15.
- Evidence-linked conclusions that distinguish facts, assumptions, and unresolved judgments. Weight: 10.
Score each dimension from zero to two. Weighted score = sum(weight * score / 2). A failed safety gate blocks acceptance.