reading-list-prioritizer

Reading List Prioritizer

This skill takes your personal reading list (a local markdown file with article titles and/or URLs) and returns a structured weekly reading plan: items scored by relevance to your role, grouped by topic, with a "Read this week" shortlist at the top.

Input:

• A markdown file (.md) with your reading list — one item per line, as [Title](URL) links, bare URLs, or plain text titles
• Your role or current focus area (stated in 1–2 sentences in your message)

Output:

• Markdown response with a prioritized "Read this week" shortlist + full list grouped by topic, each item labeled [High] / [Med] / [Low]

---

Language Detection

Detect the user's language from their message:

• If Russian (or contains Cyrillic): respond in Russian
• If English (or other Latin-script language): respond in English
• If ambiguous: respond in the language of the trigger phrase used

---

Instructions

Step 1: Validate Input

1. Check that a reading list file has been provided (uploaded or pasted)

   • If no file and no pasted list: stop and return: "Please provide your reading list as a .md file or paste the list directly."
   • Accepted formats: .md file, pasted markdown list, plain text list

2. Check that the user's role or focus area is stated in their message

   • If missing: ask for it before proceeding — "What's your current role or focus area? (e.g., 'PM focused on AI product strategy')"
   • Do not guess or assume a generic role — prioritization depends on it

3. Validate file content: must contain at least 3 parseable items

   • If fewer than 3 items: proceed but note "Short list — showing all items with priority labels"
   • If file is empty or contains no recognizable items: stop with message "The reading list appears empty. Please check the file format."

Step 2: Parse the Reading List

1. Read each line of the list and extract:
   • Title: from [Title](URL) format, or plain text line, or domain-derived label if only a bare URL
   • URL/domain: extract domain from URL if present (e.g., nngroup.com, hbr.org, stratechery.com)
   • Tags or notes: if the user appended a tag or comment after the link (e.g., [Article](url) #ux), extract them