Skillget

no-more-fomo

agent
Security Audit
Warn
Health Warn
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 6 GitHub stars
Code Warn
  • fs module — File system access in scripts/render.js
Permissions Pass
  • Permissions — No dangerous permissions requested
Purpose
This agent aggregates daily AI news from sources like Twitter, arxiv, HackerNews, and podcasts into a single, summarized daily digest. It delivers the output directly to your local machine as Markdown and styled HTML files.

Security Assessment
Overall risk: Low. The tool does not request dangerous permissions and contains no hardcoded secrets. It makes network requests to fetch RSS feeds and public web content, which is standard for its purpose. The scanner flagged file system access in `scripts/render.js`, but writing to a local `~/no-more-fomo/` folder is the expected behavior for generating the digests. It does not access sensitive user data or execute hidden shell commands.

Quality Assessment
The project is actively maintained, with its most recent push occurring today. It uses the permissive MIT license and includes clear documentation with usage examples. The main drawback is low community visibility. With only 6 GitHub stars, the codebase has not been widely vetted by a large audience. Developers should expect to rely primarily on the original author for ongoing support and bug fixes.

Verdict
Safe to use, though standard caution is advised given its low community adoption.
SUMMARY

Never miss what matters in AI. Daily digest from 16 Twitter KOLs, 3 AI labs, 4 podcasts, arxiv topic search, HuggingFace trending papers, and HackerNews — one command, every morning.

README.md

no-more-fomo

You wake up. 47 new arxiv papers. 200 AI tweets. 3 podcast episodes. A new model just dropped.

You're already behind.

Or are you?

One command. Every morning. Never miss what matters.

/no-more-fomo turns the firehose of AI news into a single enriched daily digest — with real summaries, real links, and real context. Not just headlines.

npx skills add freemty/no-more-fomo

See a real example digest to know exactly what you get.

What makes this different

Most "AI news" tools give you a list of links. You still have to click each one, read it, and decide if it matters.

/no-more-fomo does the reading for you:

  • Papers come with 2-3 sentence summaries from the actual arxiv abstract
  • Podcasts come with structured summaries: TLDR, chapter breakdown, and speaker-attributed quotes extracted from transcripts
  • HN threads come with context on why the community cares
  • Everything has real links — arxiv, GitHub, HuggingFace, not t.co
  • Every digest also outputs a styled HTML version with three-view layout, dark mode, and date navigation

Sources

Category What Count
Twitter KOLs @karpathy, @_akhaliq, @emollick, @simonw, @swyx, and 11 more 16 default
AI Lab Blogs DeepMind, Anthropic, OpenAI 3
Podcasts No Priors, Latent Space, Dwarkesh, Training Data (Sequoia) 4
arxiv Papers Topic-filtered search (default: AI agent + LLM) per topic
HuggingFace Daily Community-upvoted trending papers top 20
HackerNews AI agent + broader AI/LLM stories 2 queries
Company accounts Cursor, xAI, Windsurf, Cognition, Replit, HuggingFace 6 (with --full)

Usage

/no-more-fomo                    # Daily digest — all default sources
/no-more-fomo --full             # Also include company/product accounts
/no-more-fomo --transcripts      # (deprecated — Phase 2 does this automatically)
/no-more-fomo @someone           # Add any Twitter handle
/no-more-fomo --hn-only          # Just HackerNews
/no-more-fomo --quick           # Quick digest — skip deep processing
/no-more-fomo --no-html          # Markdown only, skip HTML output
/no-more-fomo --podcasts-only    # Just podcast feeds

Saves to ~/no-more-fomo/YYYY-MM-DD.md.

Customize your sources

Create ~/.no-more-fomo/config.yaml to add/remove sources. You only specify what you want to change — everything else keeps the defaults.

twitter:
  add:
    - handle: "@elonmusk"
      count: 15
    - handle: "@sama"
      count: 15
  remove:
    - "@ylecun"

podcasts:
  add:
    - name: "Lex Fridman"
      rss: "https://lexfridman.com/feed/podcast/"
      transcript: youtube

papers:
  topics:
    - query: "world model"
      categories: ["cs.AI", "cs.CV"]
    - query: "reinforcement learning"
      categories: ["cs.LG"]
  hf_daily: true
  min_hf_upvotes: 5

hn:
  extra_queries:
    - "robotics"

discovery:
  enabled: true
  max_per_topic: 3

topic_search:
  enabled: true

language: zh   # en | zh

See SKILL.md for the full config reference.

Install

Via skills.sh (recommended)

npx skills add freemty/no-more-fomo

Works with Claude Code, Cursor, Codex, Windsurf, and 15+ other agents.

Manual

git clone https://github.com/freemty/no-more-fomo.git ~/.claude/skills/no-more-fomo

Prerequisites

  • xreach (npm i -g xreach-cli) — Twitter/X data
  • yt-dlp — podcast transcripts (optional)
  • baoyu-youtube-transcript (optional) — enhanced podcast transcript processing
  • bun — runtime for youtube-transcript (optional)
  • curl — RSS feeds and HN API (standard on all systems)

Schedule it

Claude Code Cloud (recommended)

Go to preview.claude.ai/code → Scheduled → + New scheduled task:

  • Name: no-more-fomo
  • Prompt: run /no-more-fomo and save the digest
  • Frequency: Daily, 09:00 AM

Runs on cloud infra — no local machine needed.

Local fallback (crontab)

0 9 * * * claude -p "run /no-more-fomo and save the digest" --allowedTools "Bash,Read,Write,Glob" --output-format stream-json >> /tmp/no-more-fomo.log 2>&1

How it works

  1. Fetch — pulls from all sources in parallel (16+ Twitter accounts, 3 blogs, 4 podcasts, HN)
  2. Parse — extracts expanded URLs from tweet entities (never outputs t.co links)
  3. Filter — last 24h for Twitter/HN, last 7 days for blogs/podcasts, quality thresholds
  4. Enrich — fetches arxiv abstracts, GitHub descriptions, podcast transcripts
  5. Categorize — Papers, Models, Tools, Agents, Lab Updates, Podcasts, HN, Industry
  6. Output — structured markdown with summaries, links, and engagement metrics
  7. Render — generates styled HTML from template with three-view layout (newspaper/sidebar/grid)

More from freemty

  • claude-code-best-practices — Workflow advisor skill — recommends the right skill, agent, or workflow for any task
  • labmate — Research harness for Claude Code
  • cc-research-playbook — AI Research with Claude Code — slides, demos, and reference materials
  • ai-dotfiles — Sync AI CLI configs across machines with automatic API key redaction

License

MIT

Reviews (0)

No results found