SnipLink — The ADHD-Friendly URL Saver

Who It's For

You found something cool. You want it saved now — before you forget, before you lose the tab, before the momentum dies.

SnipLink saves it instantly: title, description, category, tags, social links. Done in seconds. No multi-step forms, no "where should I put this" paralysis.

Trigger

Use this skill when:

1. User shares a URL and wants it saved — "save this", "remember this", "add to my stash"
2. User shares an X/Twitter link — detect tweet content, extract the real target, and save it
3. User asks "what tools do I have for X" or wants to search their saved links
4. User is brainstorming and needs quick tool suggestions by tag or category

Zero friction. One URL in, clean record out. Confirm once, forget about it.

Workflow

1. Saving a Link (One-Shot)

Step 1: Detect source type

• X/Twitter URL (x.com, twitter.com) → go to X/Twitter Pipeline (see below)
• GitHub.com repo URL → use gh api (structured data, no scraping)
• All other URLs → use web_fetch

Step 2: Extract info

For GitHub repos, extract:

• Repo name, description, primary language, star count, license, topics/tags, last updated, owner

For all other pages:

• web_fetch → title, meta description, pricing, features
• Skip if behind login, paywall, or CAPTCHA

Step 3: Auto-categorize

• AI/ML — AI services, LLMs, machine learning
• Development — Coding tools, APIs, frameworks, testing
• Productivity — Task management, notes, workflow automation
• Marketing — SEO, social media, ads, content
• Design — Graphics, UI/UX, video, prototyping
• Finance — Billing, accounting, payments
• Communication — Messaging, email, calls, CRM
• Data — Analytics, databases, visualization
• Other — Anything else

Step 4: Auto-generate tags

• Extract from description, GitHub topics, or page keywords
• Common: free, paid, api, no-code, open-source, mobile, cloud, etc.

Step 5: Social media lookup

• web_search tool name + "LinkedIn" / "Twitter"
• Store URLs if found

Step 6: Present for approval (MANDATORY)

• Show the user a clean summary of what was extracted:
  • Title, description, category, tags, price
  • Source (direct URL / tweet / GitHub)
• Ask: "Save this? (yes / no / edit)"
• If user says no → discard, ask if they want to modify
• If user says edit → let them adjust fields before saving
• If user says yes → save to database
• NEVER save without user seeing the extracted data first

X/Twitter Pipeline

Trigger: User shares an x.com or twitter.com URL and wants it saved.