Center for Federal Accountability › Reports

CFA comment drafting workflow — overview

End-to-end Federal Register ingestion, analysis, and CFA-style comment draft generation at a glance.

Purpose

The FederalRegister project supports an end-to-end pipeline around Federal Register content (including proposed rules and SORNs): ingest to PostgreSQL, enrich, run AI-assisted analysis, and generate CFA-style public comment drafts where applicable.

Core goals

Regulatory tracking: Identify rules open for comment and closing dates.
Legal and compliance analysis: Summaries and audits (e.g. SORN/PIA) stored as structured analysis records.
Comment drafting: High-quality draft comments aligned with CFA voice and procedure.
Automation: Scriptable steps with outputs under output/ and persisted results in the database.

Major components (conceptual)

Area	Role
Ingest and sync	Federal Register API → documents, agencies, full text backfill
Open comment sync	Focus on documents with active comment windows
AI analysis	Summaries, audits, comment drafts via configured LLM providers
Reports	Markdown/JSON aggregates including open-comment and accountability exports

Relationship to CFA

Downstream artifacts (JSON reports) are copied into the CFA frontend repository for static publication on federalaccountability.com. This Documentation.AI site documents how those artifacts are produced; it does not replace the FederalRegister database or runners.

Source

Condensed from FederalRegister/docs/cfa_comment_workflow.md. For file-level references (scripts, modules), use the FederalRegister repository directly.