PolicyWatcher
Loading verified public evidence
The page will appear when its current evidence state is available.
PolicyWatcher
The page will appear when its current evidence state is available.
This page describes what the current platform exposes: public dashboard contracts, change views, canonical analysis fields, provenance-aware exports, dataset QA checks, retrieval evidence, source suspension logic, Trust & Quality evidence and admin tools.
The public and admin pages are built around companies, policies, policy changes, check logs, region impacts, subscribers and QA results. The panels below map those records to the visible features.
The platform displays configured companies, source URLs, jurisdiction labels, policy categories and Dataset QA status.
When change rows exist, canonical share links preserve the committed dashboard view and browser history restores valid filter state.
Canonical KPI semantics, atomic region and audience selection, and an exact-value inspector keep visual drill-downs consistent with the evidence table.
The admin area includes five-stage Bulk Source Onboarding, Dataset QA decisions, Review Log, KPI audit, companies, cron status, VPS services and database controls.
The animated SVGs below are static representations of existing product views. They describe UI modules and the database fields they use.
Separates evidence-gated provider policy changes from sanitized source-retrieval and publication-state transitions.
Plots evidence-gated score records, distinguishes observation sequence from the real snapshot version, and includes an accessible table.
Selecting a cell commits region and audience together, while the bilingual summary and exact-value table keep missing cells as Not assessed.
Opens original and normalized values by canonical KPI, makes missing and tie states explicit, and keeps the ordinal comparison separate from compliance or performance claims.
Uses one field-specific vocabulary and concern order across normalization, audit and the public matrix.
Generates a report from the selected policy and its available analysis fields.
The admin QA layer checks source fit, retrieval evidence, public evidence gates, seeded-record boundaries, hash consistency, check-log completeness, timestamp integrity, archive timestamps, KPI coverage, regional impact coverage and subscriber hygiene. Review decisions are persisted and written to an append-only log.
The admin area includes login, metrics, company and policy management, dataset QA, KPI audit, explainability, cron status, database inspection, encrypted backup utilities, renderer-aware retrieval evidence and policy telemetry.
CSV/TSV intake, official-source review, private baseline, QA gate and publish/hold/reject decision
Source fit, public evidence gates, check logs, hash consistency and analysis field checks
KPI value distribution and explanation coverage by company
Companies, industries, policy URLs, policy types and jurisdictions
Batch scans, company targeting, live strategy evidence and source suspension results
Renderer health checks, bearer-auth smoke test and redirect/SSRF boundary checks
Database inspection for records used by public views
Admin login and access events with IP, user-agent, path and outcome
Encrypted export and decrypt-preview utility for database content
Read-only review access for selected admin views
Session HMAC, rate limits, protected APIs and token hygiene
Public quality page for CI, CodeQL, OpenSSF, targeted reliability coverage and live header reports
This is a functional map of routes and records used by the app. It separates what is configured in the dataset from what is shown in public pages.
Validated CSV/TSV candidate enters protected admin review and remains private
Official-source decision, private first baseline, QA gate and explicit publish/hold/reject are persisted
Company, policy URL, type, jurisdiction and configured status
Direct fetch, HTTP/2, VPS renderer, Wayback and Common Crawl are tried with recorded outcomes
Minimum text, host drift, path drift, extraction cap and content-shape checks
Status, source, HTTP status, final URL, hash, text length and archive timestamp are recorded
Only source-verified evidence is available to public routes and reports
Seeded inventory is replaced only when no source evidence or public baseline already exists
Diff, AI summary, KPI fields and regional impacts are generated only after a valid baseline comparison
Dataset QA decisions, admin review log and suspension alerts document human and system actions
Public evidence page links repository checks, security scans and external report surfaces
Dashboard, timeline, signals board, share, embed, report and digest routes read only gated data
The Trust page presents workflow status, CodeQL, OpenSSF Scorecard, targeted reliability coverage, Sonar readiness, Codecov readiness and live HTTP header report links. The Policy Signals Board turns source evidence into an inspectable operational ranking. These checks describe build health, retrieval quality, source status, and public evidence readiness.
The public interface explains its methodology and exposes available analysis outputs with confidence boundaries. The admin interface adds tools to inspect and improve the dataset behind those outputs.
