tenbrains

Analyzes X posts and YouTube transcripts into structured research persisted in a local SQLite database.

Updated Jun 20, 2026
One-click install
npx skills add https://github.com/moeghashim/tenbrains-arch --skill tenbrains-moeghashim
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: tenbrains
Source: https://github.com/moeghashim/tenbrains-arch/tree/main/skill
Command: npx skills add https://github.com/moeghashim/tenbrains-arch --skill tenbrains-moeghashim

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Researching X/Twitter posts and YouTube videos usually means scattered notes, lost links, and no way to recall past findings. This Skill runs the local tenbrains CLI to analyze posts and transcripts, summarize followed accounts, manage bookmarks and suggestions, and build 7-day learning tracks — with every outcome persisted to a local SQLite database and returned as parseable JSON. ## Core Features & Use Cases - Post and video analysis: Analyze a tweet by URL (free via oEmbed, no API key) or a YouTube video's captions into topic, summary, intent, and novel concepts, with optional --learn to generate a 7-day Feynman learning track with lessons and quizzes. - Account research and recall: Summarize a followed account's recent posts into takeaways, rank suggestions from saved signals, and search past analyses with full-text search scoped by learning objective. - Learning objectives and daily loop: Create first-class learn/follow objectives, tag records to them, and coach through daily lessons and graded quiz answers via learn today, learn quiz, and learn answer. - Use Case: A user pastes a tweet URL and asks to analyze it and turn it into a study plan; the Skill runs tenbrains analyze --url ... --learn, then uses learn today each day to walk them through the materialized lessons and quizzes. ## Quick Start Ask the assistant to analyze a tweet or YouTube video by URL with tenbrains, for example: "Analyze this tweet https://x.com/jack/status/20 and generate a learning track from it."

Frequently Asked Questions about tenbrains

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I analyze a tweet without an X API key?

Run tenbrains analyze with the tweet URL, for example tenbrains analyze --url "https://x.com/jack/status/20". Single tweets are fetched free through X's oEmbed endpoint, so no Bearer token is required; the token is only needed for fetching full timelines.

How do I analyze a YouTube video transcript from the command line?

Use tenbrains analyze --url with the video URL to fetch public captions, optionally adding --summarize and --learn. If captions are unavailable, supply a transcript manually with --transcript @captions.txt; audio download and transcription are not supported.

Can I use tenbrains offline without an API key?

Yes, add --provider mock to any analyzing or materializing command for deterministic, offline output with no network or credentials. The mock provider uses keyword extraction and templates lessons and quizzes, so it is suited for testing the data flow rather than real analysis.

Where does tenbrains store analysis results and credentials?

All posts, analyses, takeaways, bookmarks, suggestions, tracks, and objectives are stored in one local SQLite file, configurable with --db. Provider keys live in ~/.config/tenbrains/config.json at mode 0600, set via tenbrains setup; no environment variables or .env files are read.

What are the limitations of the tenbrains CLI?

It cannot post, reply, or DM on X, and it only covers X and YouTube content. Fetching account timelines requires a paid X API tier, YouTube support is caption-only, and the mock provider does not perform real model judgement.