# Adviserry

> Adviserry reads every newsletter and YouTube video your experts publish, weighs it against what you are building, and drafts the specific actions worth taking this week, with every suggestion cited to its source.

Adviserry helps solo founders, consultants, and professionals turn the newsletters and YouTube channels they already subscribe to into a working advisory panel. Users connect Gmail and add YouTube channels, content is auto-detected, extracted, and embedded, then it's synthesized into drafted actions, searchable via dashboard, AI chat, or a personal MCP server (compatible with Claude Desktop, Claude Code, Cursor, Gemini CLI, and other MCP clients that support bearer-token servers).

## Product

- [Homepage](https://adviserry.com): Landing page with interactive demo, pricing, and product overview
- [Full Product Documentation](https://adviserry.com/llms-full.txt): Comprehensive product details, features, pricing, FAQ, and technical information
- [MCP server overview](https://adviserry.com/mcp): What the MCP server does and how to connect it
- [MCP setup docs](https://adviserry.com/docs/mcp): Step-by-step Claude Desktop / Cursor / Claude Code connection guide
- [Compare](https://adviserry.com/compare): How Adviserry differs from Feedly, Readwise Reader, and Meco
- [Blog](https://adviserry.com/blog): Articles on newsletter overload, AI reading workflows, and product updates

## Core Capabilities

- **Auto-Ingestion**: Connect Gmail once; newsletters sync automatically (hourly). Add YouTube channels; new long-form videos are transcribed and indexed as they publish (every 6 hours).
- **Action Synthesis**: Adviserry weighs new content against what you told it you're working on and drafts the specific, ready-to-use actions worth taking. Core gets one pass a week, on Fridays; Pro gets a draft pass on weekdays plus a deeper Friday pass. Every action cites the source it came from.
- **Panels**: Group creators into topic-focused panels (e.g., "Marketing," "SaaS," "AI"). 2 panels on Core, up to 5 on Pro, and up to 10 on Team, with 10 sources each.
- **AI Chat** (Pro and Team): Ask questions grounded in your creators' actual newsletters and video transcripts, combined with your uploaded documents, for answers specific to your situation.
- **MCP Server**: Connect your panels, drafted actions, and context to Claude Desktop, Claude Code, Cursor, Gemini CLI, or any MCP client that supports bearer-token servers, via a personal API endpoint.
- **Context Uploads**: Upload pitch decks, strategy docs, or project briefs (PDF, DOCX, TXT, MD, RTF, and more) so answers and actions reference your specific situation.
- **Daily Summaries**: Automated email digests of what's new across your panels, delivered at your preferred time.

## Content Sources

- Email newsletters via Gmail (read-only OAuth, auto-detected senders) or forwarded from any inbox to a private Adviserry address
- YouTube channels (transcripts extracted, long-form videos only)
- User-uploaded documents (PDF, DOCX, DOC, TXT, MD, RTF, PPTX, PPT, XLSX, XLS, CSV, JPG, PNG, XML, up to 10 MB)

## Pricing

- **Adviserry Core**: $14.99/month with a 14-day free trial: auto-ingested Gmail newsletters and YouTube channels, 2 panels (10 sources each), Action Synthesis once a week (Fridays), MCP for Claude Desktop, Cursor, and Claude Code, context uploads, up to 5 proactive insights per panel per day, and daily summaries. No AI chat or memory, and insights are not included in the daily email
- **Adviserry Pro**: $24.99/month with a 14-day free trial: everything in Core, plus up to 5 panels, Action Synthesis every weekday with a deeper Friday pass, AI chat across your archive, memory, unlimited proactive insights, and insights in the daily email
- **Adviserry Team**: $99.99/month, 5 seats included (owner counts as one), extra seats $15/month each, one shared archive across the team
- Cancel anytime before the trial ends and pay nothing. There is no free plan.

## Technical

- Built with Next.js, Supabase (Postgres + Auth), OpenAI (gpt-4o-mini for chat/summaries, text-embedding-3-small for semantic search), and Anthropic Claude (Sonnet/Opus for Action Synthesis)
- MCP server implements JSON-RPC 2.0 protocol with 13 tools: search_newsletters, list_panels, get_recent_issues, get_newsletter_summary, get_topics, search_context, get_insights, search_actions, get_actions, get_action, update_action_status, get_user_context, update_context
- MCP rate limit: 500 requests per rolling hour per user
- Content extraction uses 3-layer fallback: Readability, Platform Parser (Substack, Beehiiv, ConvertKit, Mailchimp, Ghost), LLM Fallback
- Vector search via pgvector (cosine similarity, 1536-dimension embeddings), both whole-issue and chunk-level
- All OAuth tokens encrypted with AES-256-GCM; row-level security on all data

## Links

- [Start your free trial](https://adviserry.com/signup)
- [Log In](https://adviserry.com/login)
- [Newsletter Stack Audit](https://adviserry.com/audit): Free tool that scores your current newsletter subscriptions
- [Operator Archetype Quiz](https://adviserry.com/quiz): Free quiz on how you consume expert content
- [Terms of Service](https://adviserry.com/terms#terms-of-service)
- [Privacy Policy](https://adviserry.com/terms#privacy-policy)
- [Contact](mailto:support@cassontech.com)