Changelog

All notable changes to Context Memo will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[Unreleased]

Changed

• Memo type routing optimized for signal generation - Scan-based memo generation now defaults to how_to format for all non-competitive query types. Data shows how_to memos generate 10-20x more ai_user_browse events per memo than comparison/alternative/industry types. Previously, 624 queries with funnel-stage types (top_funnel, mid_funnel, bottom_funnel) fell through with no memo generated at all.

Added

• Screenshot Freshness System - Automated screenshot capture and staleness tracking for the image library
  • Mark images as stale/current from the library UI
  • Stale images excluded from memo generation
  • New screenshot/capture automation captures product pages via ScreenshotOne API
  • Auto-triggered after deploys that change UI files
• Citations Tab Overhaul - Major upgrade to the Citations page
  • Citation count now shows in the tab label (e.g., CITATIONS (47))
  • New My Content view showing which of your memos are being cited by AI models, citation counts, and which prompts cite them
  • Brand domain rank callout in Domains view (e.g., "You rank #11 of 45 domains")
  • Content match indicators: "Covered" and "Gap" badges on cited URLs
  • Generate Memo CTA for top cited URLs with no matching content

Planned

• Additional memo templates (best-of, what-is)

[0.20.0] - 2026-02-06

Added

• Content Coverage Audit - New COVERAGE tab on brand page that maps the complete content topic universe
  • Crawls brand's sitemap.xml to build a full inventory of existing website content
  • AI generates 40-150 specific content topics the brand needs (comparisons, alternatives, how-tos, industry guides, definitions, use cases)
  • Each topic is scored by priority (0-100) based on competitor density, content type, and funnel stage
  • Topics are matched against existing content (website pages + memos) and marked as covered/partial/gap
  • Coverage score with circular progress indicator and category breakdown
  • "Top Priority Gaps" section with stack-ranked highest-impact missing content
  • One-click "Generate" button per gap topic to create a memo
  • Batch generate top 10 gaps at once
  • Continuous monitoring: auto-adds comparison/alternative topics when new competitors are discovered in scans
• Site Content Inventory utility (lib/utils/site-inventory.ts) - Fetches sitemap.xml, batch-classifies URLs by content type, and deep-reads key pages for quality assessment
• Topic Universe database table and Inngest functions for generation and refresh
• Coverage API endpoint at /api/brands/[brandId]/coverage

[0.19.5] - 2026-02-06

Fixed

• Memo Sources: Sources section now shows real clickable URLs instead of AI-hallucinated descriptions
  • Was showing vague text like "Perplexity AI documentation and behavior analysis"
  • Now generates proper markdown links: [Brand Name](https://brand.com) (accessed date)
  • Fixed {{brand_domain}} placeholder that was never being replaced in generation code
  • Fixed ALTERNATIVE prompt {{competitor_name}} only replacing first occurrence
  • Database sources array now includes competitor domain for comparison/alternative memos

[0.19.4] - 2026-02-06

Added

• Daily Digest Email: Automated daily visibility summary sent via Resend at 9 AM ET
  • Per-brand visibility score with delta vs yesterday
  • Scan count, mention rate, citation rate
  • New citations won and citations lost
  • Memos generated/published count
  • Competitor content detected
  • Top competitors by mention frequency
  • Prompt coverage bar (X of Y prompts citing your brand)
  • Streak milestones (queries with 5+ day citation streaks)
  • Skips tenants with no activity (no spam on quiet days)
  • Clean, responsive HTML email template

[0.19.3] - 2026-02-06

Changed

• Memo-First Branding: Rebranded all "content" references to "memos" across the product UI
  • Landing page: "Content Intelligence" → "Competitive Watch", "Auto-Generate Content" → "Generate Memos", "Popular Content" → "Featured Memos"
  • New trust section: "Not Another Blog Post" — positioning memos as structured, brand-verified, built for AI
  • Settings: "Content Settings" → "Memo Settings" with updated descriptions
  • Updated copy across activity detail, memo editor, competitor list, entity list, and brand actions
  • Left "competitor content" references untouched (refers to their blog posts, not our memos)

[0.19.2] - 2026-02-06

Added

• Popular Content Section on Homepage: New section showcasing featured memos to drive traffic to top content
  • Displays up to 6 featured memos with type badges, titles, and descriptions
  • "View All Content" button links to /memos index
  • Curated via featured boolean column on memos table
  • sort_order column controls display order

Changed

• Database: Added featured (boolean) and sort_order (integer) columns to memos table

[0.19.1] - 2026-02-05

Fixed

• Subdomain Memo Pages 404: Fixed all memo pages returning 404 on brand subdomains (e.g., checkitnet.contextmemo.com/how/...)
  • Root cause: Supabase query had invalid relational join (reviewed_by:reviewed_by(...)) that failed silently
  • Also fixed metadata query missing brand_id filter
  • All memo URLs now resolve correctly via subdomain routing

[0.19.0] - 2026-02-05

Added

• Public Content Routes for Context Memo Brand: SEO-optimized URL structure for publishing memos on contextmemo.com

  • /compare/[slug] - Comparison memos (vs competitors)
  • /alternatives/[slug] - Alternative memos (alternatives to X)
  • /guides/[slug] - How-to guides and tutorials
  • /tools/[slug] - Tools, resources, and best-of lists
  • /for/[slug] - Industry-specific content
  • Index pages for each content type with memo listings
  • Backwards compatible with existing slug formats

• Shared Memo Rendering Utility: lib/memo/render.tsx

  • Reusable memo page component
  • Markdown processing and HTML rendering
  • AI-optimized Schema.org structured data
  • Review status badges and provenance tracking

[0.18.0] - 2026-02-05

Added

• Corporate Positioning Framework: Comprehensive 32-field strategic messaging framework

  • 8 sections: Mission & Vision, Target Markets, Value Proposition, Key Differentiators, Messaging Pillars, Elevator Pitches, Objection Handling, Competitive Stance
  • Inline edit capability for each section with save/cancel
  • Progress tracking showing completion percentage
  • Collapsible sections for easy navigation

• Corporate Positioning API: New endpoint /api/brands/[brandId]/corporate-positioning

  • PATCH for section-level updates
  • PUT for full positioning replacement
  • GET for retrieving positioning data

• Retroactive Data Population: All existing brands now have corporate positioning data

  • AI-generated positioning based on website content and web research
  • Includes mission statements, buyer personas, differentiators, pitches, and competitive intelligence

[0.17.0] - 2026-02-04

Added

• QFO (Query Fan-Out) Tab: New tab to analyze how LLMs expand prompts

  • Generate 6-8 fan-out sub-queries from any prompt using AI
  • Scan each sub-query with Perplexity to check brand coverage
  • Visual coverage summary (cited/mentioned/gaps/competitor wins)
  • One-click content cluster generation for gap queries
  • Based on the QFO SEO strategy for AI visibility

• MAP (Entity Discovery) Tab: Interactive ReactFlow visualization

  • Radial graph with your brand at center
  • Entities discovered from AI scan results (competitors_mentioned, citations)
  • Color-coded by type: Competitor (red), Resource (green), Aggregator (amber), Publisher (purple)
  • Node size reflects mention frequency
  • Animated edges highlight competitor wins (they're cited, you're not)
  • Click entities for detailed stats and analysis actions
  • Mini-map for navigation
  • Stats summary row

• New API Endpoints:

  • /api/brands/[brandId]/qfo - Generate fan-out queries, scan coverage, generate clusters
  • /api/brands/[brandId]/entities - Extract and classify entities from scan results

Dependencies

• Added @xyflow/react (ReactFlow v12) for interactive graph visualization

[0.16.0] - 2026-02-04

Added

• HubSpot Landing Page: New /hubspot page for marketing to HubSpot users
  • Focused value proposition: "Fill the gaps in your content strategy"
  • 4-step workflow visualization with product screenshots
  • Consumption-based pricing calculator with interactive sliders
  • Beta signup with 50% discount offer

[0.13.0] - 2026-02-02

Added

• RSS Feed Monitoring: Proper RSS/Atom feed parsing with rss-parser library
  • Auto-discovers feeds from competitor sites via common paths + HTML autodiscovery
  • Persistent competitor_feeds table tracks discovered and manual feeds
  • Feed health monitoring (last success, failures, errors)
• Historical Content Backfill: One-click button to import up to 2 years of competitor content
• Manual Feed Management: Add custom RSS/Atom URLs per competitor in the UI
• Enhanced Content Tracking: Published date, author, word count stored for all competitor articles
• Improved Response Generation: New prompt creates unique, differentiated, LONGER content (1000-1500 words)
  • Emphasizes deeper analysis, unique angles, actionable frameworks
  • Uses full competitor content for better differentiation
  • Creates more specific, compelling titles

Changed

• Competitor content scan now uses proper RSS parser instead of regex
• Response memos target 1000-1500 words (up from 600-900)
• Feeds are discovered once and persisted for efficient re-scans

[0.12.0] - 2026-02-02

Added

• Automated memo generation with daily caps: System now enforces memos_per_month / 30 daily limit
• Competitor-linked queries: Query generation now creates proper alternative and versus queries with competitor IDs linked
• sanitizeSlug helper: Removes special characters (?!., etc.) from memo URLs
• Live page links: Published memos now show clickable links to live pages with external link icon
• Full URL display: Dashboard shows subdomain.contextmemo.com/slug instead of just /slug

Changed

• Query generation prompt now requires 5-8 alternative + 5-8 versus queries with competitor names
• Memo generation now handles persona_based, problem_solution, best_of query types
• Improved memo template styling with better typography, spacing, and visual hierarchy
• Scan-run orders gap queries by priority (high-value first)

Fixed

• Memos no longer generate with trailing ? in URLs
• Query types persona_based and problem_solution now trigger memo generation (previously ignored)
• best_of queries now properly map to industry memos

[0.11.0] - 2026-02-01

Added

• Activity Feed: Collapsible sidebar showing all platform activity in real-time
  • Aggregates activity from scans, memos, competitor discoveries, queries, AI traffic
  • Timestamped entries with links to relevant pages
  • Filter by category (Scans, Content, Discovery, Traffic, System)
  • Filter by brand (for multi-brand users)
  • Saved views - save filter combinations for quick access
  • Activity icon button in header to open feed
• Activity Detail View: Click any activity to see full context
  • "What This Means" explanations for each activity type
  • "Why It Matters" business context
  • "Next Steps" action guidance
  • Shows actual data: generated prompts, scan results, memo content
  • Explains prompt types: persona-based, intent-based, category-based
• Actionable Visibility Gaps: Scan results now show each "gap" prompt with actions
  • See exactly which prompts have 0% visibility
  • "Generate Memo" button to create content for that prompt
  • "Edit Prompt" button to refine the prompt text
  • "Disable Prompt" button to exclude irrelevant prompts
  • Shows visibility %, prompt type, and persona for each gap
• New API endpoints: /api/activity, /api/activity/views, /api/activity/detail
• Activity types and utility logging functions for future Inngest integration
• SQL migration for activity_log and activity_saved_views tables

Changed

• Refactored dashboard header into client component for activity feed integration

Fixed

• Failed memo detection: Memos now validate AI responses and reject error messages
• Regenerate button: Added to memo edit page to fix broken memos
• Warning banner: Shows when memo content looks like an AI error with action buttons

[0.10.1] - 2026-02-01

Fixed

• TypeScript type errors for AI providers in scan-run.ts (OpenAIProvider vs LanguageModelV3)
• Stripe client proxy type casting for strict TypeScript checking

[0.10.0] - 2026-02-01

Added

• Google AI Overviews Integration: Check brand visibility in Google's AI-generated summaries
  • Manual trigger button in dashboard header
  • Automated weekly scans (Mon + Thu) via SerpAPI
  • Detects brand mentions, organic rankings, and source citations
• AI Traffic Attribution: Track visits from AI platforms (ChatGPT, Perplexity, Claude, Gemini, etc.)
  • Auto-tracking on memo pages
  • Dashboard view with source breakdown and top pages
  • Embed code for external website tracking
• Data Export: CSV and JSON exports for scans, prompts, memos, competitors, visibility history
• Multi-seat Teams Foundation: Organizations with roles (owner, admin, member, viewer)
  • Invite flow with email tokens
  • Team management UI
• Stripe Billing Foundation: Checkout, customer portal, webhooks, plan definitions
  • Starter, Growth, Enterprise tiers
  • Usage tracking foundation
• 3 New AI Models: DeepSeek V3, Qwen 2.5 72B, Grok 2 (9 total models now)

Changed

• Upgraded to 9-model AI scanning coverage
• Weekly automation now includes AI Overview scans (quota-conscious: Mon + Thu only)

[0.9.0] - 2026-02-01

Added

• Terminal-style onboarding experience: New brands see an engaging step-by-step setup with live terminal output
  • Real-time progress messages showing what's happening ("Connecting to Jina Reader API...", "Analyzing products...", etc.)
  • Dark terminal UI with timestamps and blinking cursor
  • Polls backend for actual completion, shows summary when done
  • Auto-refreshes to proceed to next step
• Cleaner onboarding flow: Step-by-step card with progress indicators instead of confusing multiple buttons
• Status check API: New check_status action for polling onboarding progress

Changed

• Removed confusing buttons: Simplified header by removing "Discovery Scan" button during onboarding
• Metrics only show after scans: No more fake 0% visibility scores for new brands
• "Ready to scan" prompt: After onboarding completes, shows clear prompt to run first scan instead of empty metrics

Fixed

• New brands no longer see meaningless metrics before any scans have run

[0.8.0] - 2026-02-01

Added

• Google Search Console Integration: Full OAuth 2.0 flow for connecting GSC accounts
  • Connect button in brand settings
  • Automatic token refresh handling
  • Weekly auto-sync (Sundays 9 AM UTC)
  • Shows Google AI Overviews pathway data
• Loading skeleton UIs: Layout-matching skeleton loaders for all dashboard pages
• Instant click feedback: Button press animation (active:scale-[0.97])

Changed

• Improved loading states across dashboard, brand page, settings, and auth pages

[0.7.0] - 2026-02-01

Added

• Bing Webmaster Integration: Search console showing which Bing queries drive traffic
  • Settings panel for API key configuration
  • Weekly auto-sync (Sundays 9 AM UTC)
  • Query correlation with AI prompts
  • "Opportunities" section for queries to target
• New search_console_stats database table

[0.6.0] - 2026-02-01

Added

• Competitive Intelligence Dashboard
  • Share-of-voice analysis across all scans
  • Win/tie/loss tracking vs competitors per query
  • "Queries to improve" section (where competitors beat brand)
  • Visual progress bars for share of voice metrics

[0.5.0] - 2026-02-01

Added

• Competitor Content Intelligence
  • Daily scan of competitor blogs/content
  • AI classification (filters press releases, feature announcements)
  • Auto-generates response articles with brand's tone
  • Auto-publishes to resources page
• New competitor_content table
• New memo type: response

[0.4.0] - 2026-02-01

Added

• Persona-based Prompt System
  • 6 persona types: B2B Marketer, Developer, Product Leader, Enterprise Buyer, SMB Owner, Student
  • Personas auto-extracted from brand website analysis
  • Prompts generated only for brand's relevant personas
• OpenRouter Multi-model Scanning
  • GPT-4o, Claude, Gemini 2.0 Flash, Llama 3.1 70B, Mistral Large, Perplexity Sonar
  • Visibility chart updated to show all 6 models

Changed

• Renamed "queries" to "prompts" throughout UI

[0.3.0] - 2026-02-01

Added

• External Credibility Signals
  • /about/editorial guidelines page
  • /ai.txt for AI crawler permissions
  • Enhanced Schema.org with sameAs links (LinkedIn, Crunchbase, Wikipedia)
  • Social links support in BrandContext
• Automated Backlinking System
  • Auto-runs after memo generation
  • Daily refresh at 7 AM UTC
  • Injects contextual links + "Related Reading" section

[0.2.0] - 2026-01-31

Added

• High-intent Query Generation: Intent-based queries from homepage content with buyer signal filtering
• Discovery Scan: Tests 50+ query variations to find where brand IS being mentioned
• Brand Tone System: Comprehensive tone controls for memo generation
  • Personality, formality, technical level
  • Audience type, writing style, jargon usage

Fixed

• Subdomain link generation (relative paths for subdomain access)
• Subdomain routing for Vercel production (x-forwarded-host support)

[0.1.0] - 2026-01-31

Added

• Initial MVP Release
• Landing page with product explanation and early adopter pricing
• User authentication (signup/login with work email + email verification)
• Email verification flow with confirmation emails
• Brand creation with email domain verification
• Dashboard with visibility scores
• Brand detail page with stats and tabs
• Settings page for brand management

Background Jobs

• context/extract - Crawls website, extracts brand context
• competitor/discover - Identifies competitors using AI
• query/generate - Generates search queries
• scan/run - Queries AI models to check brand mentions
• memo/generate - Creates factual memos (comparison, industry, how-to, alternative)
• discovery/scan - Tests query variations for baseline discovery
• daily/run - Main scheduler for all automation

Public Pages

• /memo/[subdomain]/ - Brand memo index
• /memo/[subdomain]/[...slug] - Individual memo pages

Version History Summary

Deployment Notes

• Hosting: Vercel (connected to GitHub main branch)
• Database: Supabase (us-east-1)
• Job Queue: Inngest
• Domain: contextmemo.com (with wildcard *.contextmemo.com for subdomains)