Changelog

All notable changes to RankDisco are documented here.

---

2025-12-24

Changed

• Directory reorganization - Major cleanup of packages/api/src/ folder structure
  • Organized endpoints/ (103 files) into 15 feature-based subfolders
  • Organized lib/ (81 files) into 11 domain-based subfolders
  • Added AGENTS.md documentation files to all folders
  • Updated all import paths across the codebase

Endpoints Structure

Folder	Purpose
admin/	Administrative operations (12 files)
apps/	App Store/Google Play endpoints (21 files)
assets/	Static asset serving (3 files)
classification/	URL/domain classification (7 files)
crawl/	Web crawling endpoints (6 files)
dataforseo/	DataForSEO webhooks (3 files)
debug/	Debugging endpoints (4 files)
domains/	Domain management (4 files)
keywords/	Keyword research/tracking (5 files)
misc/	Utility endpoints (4 files)
projects/	Project management (3 files)
social/	Social media scraping (1 file)
test/	Test endpoints (16 files)
tracking/	Ranking tracking (10 files)
workflows/	Workflow triggers (2 files)

Library Structure

Folder	Purpose
classification/	Classification pipeline (15 files)
crawl/	Crawling utilities (4 files)
dataforseo/	DataForSEO API client (7 files)
domains/	Domain/URL utilities (3 files)
integrations/	Third-party clients (6 files)
keywords/	Keyword management (19 files)
parsing/	HTML parsing (4 files)
social/	Social scraping (5 files)
storage/	Storage clients (3 files)
utils/	General utilities (13 files)
workflows/	Workflow helpers (2 files)

Documentation

• Updated architecture overview with new project structure
• Each folder now has an AGENTS.md file documenting its contents

---

2025-12-23

Added

• Roadmap page - Comprehensive development roadmap organized by phases
• Social infrastructure documentation - Documented existing social scraping capabilities
• Workflow documentation - Comprehensive workflow docs with Mermaid diagrams
• Public API reference - Client-facing API documentation
• Internal API reference - Admin and operations API documentation
• Mermaid support - Enabled diagram rendering in docs site

Fixed

• Keywords data model normalization - Fixed FK constraint error in domain onboard workflow
  • Added UNIQUE constraint on keywords(keyword) to enable ON CONFLICT upserts
  • Created project_keywords VIEW as alias for keyword_subscriptions junction table
  • Updated bulk_save endpoint to use global keywords + subscription links
  • Updated recategorize endpoint to JOIN through keyword_subscriptions

Changed

• Keywords are now global (no project_id on keyword record)
• Project-keyword relationships stored in keyword_subscriptions junction table
• Same keyword can be tracked by multiple projects without duplication

Documentation

• Updated DomainOnboard workflow diagram to show full social discovery flow
• Added Brand Properties Discovery section explaining the 5-step process
• Added Social Infrastructure section with implementation status table
• Documented cascade fetch strategy for profile scraping
• Clarified what's implemented vs planned (SharedCount, official APIs)

---

2025-12-22

Fixed

• ESLint cleanup - Resolved all ESLint errors in lib/*.js files
• TypeScript errors - Fixed type errors in workflow files
• Removed unnecessary regex escape characters in classifier-llm.js

Added

• TypeScript, ESLint, and Prettier configuration to API package

---

2025-12-18

Added

• Domain onboard workflow - Complete domain intelligence gathering
  • Backlink fetching and classification
  • Keyword fetching and classification
  • Brand properties discovery
  • Social account detection and verification
  • Ranking domain properties scraping
• SERP tracking workflow - Location-aware keyword ranking tracking
• Classification workflows - URL, keyword, and domain classification pipelines
• Subscription dispatch workflow - Daily cron for recurring tracking
• App details workflow - App store metadata enrichment

Changed

• Migrated from queue-only to workflow-first architecture
• Added cost tracking for all DataForSEO API calls

---

2025-12-01

Added

• Initial release of RankFabric edge worker
• Cloudflare Workers + D1 + Queues + Workflows architecture
• DataForSEO integration for backlinks, keywords, SERP, apps
• Multi-stage classification pipelines (rules → vectorize → content → LLM)
• App store tracking for Apple and Google Play
• YouTube channel resolution
• Social account discovery from domain homepages
• Cost tracking across all external APIs

---

Migration Notes

Keywords Schema (2025-12-23)

If you have existing data, run migration 0122_normalize_keywords_table.sql:

-- Add UNIQUE constraint on keyword text
CREATE UNIQUE INDEX IF NOT EXISTS idx_keywords_keyword_unique ON keywords(keyword);

-- Create project_keywords VIEW
CREATE VIEW IF NOT EXISTS project_keywords AS
SELECT 
  id, project_id, keyword_id, channel, frequency, status,
  metadata_json, last_run_at AS last_checked_at, next_run_at AS next_check_at,
  last_status, error_message, created_at, updated_at,
  NULL AS location_code, NULL AS device, NULL AS language_code,
  NULL AS target_domain, NULL AS notes, NULL AS asset_id
FROM keyword_subscriptions;

---

Versioning

RankFabric follows Semantic Versioning:

• Major: Breaking API changes
• Minor: New features, backward compatible
• Patch: Bug fixes, backward compatible