Claude Skill

ai-news

Answers "what's new in AI", "latest on <model/company/topic>", "AI news this week", "who raised" and similar questions about recent AI models, companies, funding rounds, tools and research, using the AI Resources daily briefings (blogs.drix10.com, CC BY 4.0) with cited sources. U

LLM Mart · 0 points · 1 views 139 listing impressions 0 install-command copies
Download drix10-ai-resources-skills_ai-news-7a6371f.zip · 1 KB

Install

skills CLI npx skills add https://github.com/Drix10/ai-resources/tree/main/skills/ai-news
Claude Code claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install drix10-ai-resources@llmmart
Git git clone https://github.com/Drix10/ai-resources.git

The skills CLI installs just this skill, for any of its supported agents. Claude Code installs the whole drix10/ai-resources collection as a plugin from our marketplace. Git is the plain clone.

Skill manifest

AI news from the AI Resources briefings

The AI Resources knowledge base publishes daily briefings on AI models, companies, funding, tools and research. Each briefing groups several news items, each with key points and links to the original posts.

Pick the access path

  1. MCP server installed (tools named search_briefings, latest, get_briefing, list_categories): use those tools and skip the rest of this section.
  2. Otherwise use web fetch on the JSON API (no key, CORS open):
    • Search: https://blogs.drix10.com/api/articles?q=<words>&limit=10
      • all words must match; keep queries to 1-3 distinctive words (anthropic, series a, gemini)
      • optional: category=<slug>, since=YYYY-MM-DD, page=N
    • Latest: https://blogs.drix10.com/api/articles?limit=10 (add category=<slug> for one topic)
    • One briefing in full Markdown: https://blogs.drix10.com/api/articles?slug=<slug or briefing URL> (the item's content)
    • Categories and slugs: https://blogs.drix10.com/api/categories
    • Each list item has title, url, date, category, description, headlines (every item in the briefing) and sources ({title, url}).
  3. Plain text instead of JSON: https://blogs.drix10.com/llms/<category-slug>.txt lists a category's 50 newest briefings; https://blogs.drix10.com/llms.txt links them all.
  4. API down: read raw Markdown on GitHub, https://raw.githubusercontent.com/Drix10/ai-resources/main/<Category Name>/resources-<N>.md (category folder names with spaces as %20; a higher N is newer). List a folder with https://api.github.com/repos/Drix10/ai-resources/contents/<Category%20Name>.

How to answer

  • For "latest"/"this week" questions, use latest or since= with a date about 7 days back. For a named model, company or person, search its name first, then a broader term if nothing matches.
  • Search matches item titles and each briefing's opening text, so a match may sit in headlines rather than the title. Fetch the full briefing (slug=) before quoting details from it.
  • Report what the briefings say with their dates. Do not fill gaps from memory; say when the briefings do not cover something.
  • Cite every claim: link the briefing url, and the original post from sources when there is one.
  • End the answer with: Data: github.com/Drix10/ai-resources (CC BY 4.0).
Files (ai-resources)
  • SKILL.md 2.7 KB
    ---
    name: ai-news
    description: Answers "what's new in AI", "latest on <model/company/topic>", "AI news this week", "who raised" and similar questions about recent AI models, companies, funding rounds, tools and research, using the AI Resources daily briefings (blogs.drix10.com, CC BY 4.0) with cited sources. Use whenever the user asks for recent AI news or developments.
    ---
    
    # AI news from the AI Resources briefings
    
    The AI Resources knowledge base publishes daily briefings on AI models, companies, funding, tools and research. Each briefing groups several news items, each with key points and links to the original posts.
    
    ## Pick the access path
    
    1. **MCP server installed** (tools named `search_briefings`, `latest`, `get_briefing`, `list_categories`): use those tools and skip the rest of this section.
    2. **Otherwise use web fetch on the JSON API** (no key, CORS open):
       - Search: `https://blogs.drix10.com/api/articles?q=<words>&limit=10`
         - all words must match; keep queries to 1-3 distinctive words (`anthropic`, `series a`, `gemini`)
         - optional: `category=<slug>`, `since=YYYY-MM-DD`, `page=N`
       - Latest: `https://blogs.drix10.com/api/articles?limit=10` (add `category=<slug>` for one topic)
       - One briefing in full Markdown: `https://blogs.drix10.com/api/articles?slug=<slug or briefing URL>` (the item's `content`)
       - Categories and slugs: `https://blogs.drix10.com/api/categories`
       - Each list item has `title`, `url`, `date`, `category`, `description`, `headlines` (every item in the briefing) and `sources` (`{title, url}`).
    3. **Plain text instead of JSON**: `https://blogs.drix10.com/llms/<category-slug>.txt` lists a category's 50 newest briefings; `https://blogs.drix10.com/llms.txt` links them all.
    4. **API down**: read raw Markdown on GitHub, `https://raw.githubusercontent.com/Drix10/ai-resources/main/<Category Name>/resources-<N>.md` (category folder names with spaces as `%20`; a higher N is newer). List a folder with `https://api.github.com/repos/Drix10/ai-resources/contents/<Category%20Name>`.
    
    ## How to answer
    
    - For "latest"/"this week" questions, use `latest` or `since=` with a date about 7 days back. For a named model, company or person, search its name first, then a broader term if nothing matches.
    - Search matches item titles and each briefing's opening text, so a match may sit in `headlines` rather than the title. Fetch the full briefing (`slug=`) before quoting details from it.
    - Report what the briefings say with their dates. Do not fill gaps from memory; say when the briefings do not cover something.
    - Cite every claim: link the briefing `url`, and the original post from `sources` when there is one.
    - End the answer with: `Data: github.com/Drix10/ai-resources (CC BY 4.0)`.
    

Comments (0)

Sign in to join the conversation.

No comments yet.

Reviews (0)

No reviews yet.