skill discovery
30 articles about skill discovery in AI news
SSL: Structured Skill Language Boosts Skill Discovery MRR to 0.707
Researchers propose SSL, a three-layer typed JSON representation for AI agent skills, replacing unstructured SKILL.md prose. Using an LLM normalizer, SSL improves Skill Discovery MRR from 0.573 to 0.707 and Risk Assessment macro F1 from 0.744 to 0.787 on a newly released 6,184-skill corpus.
Stop Thinking 'Progressive Disclosure' for Claude Skills — Think
A mental model shift from 'progressive disclosure' to 'progressive discovery' makes building Claude Skills more intuitive by clarifying Claude's active role in finding what it needs.
SciSpace Evolves: From AI Research Assistant to Full Workflow Platform with 'Skills'
SciSpace is expanding beyond its core AI tools for paper discovery and writing by introducing external app integrations and customizable 'Skills,' aiming to become a true all-in-one research workflow platform rather than just a collection of features.
Ctx2Skill: Self-Play Framework Lets LMs Discover Skills Without Labels
Ctx2Skill discovers skills from context via multi-agent self-play without labels. Outputs plug into any LM, targeting manual prompt engineering bottlenecks.
Anthropic's Claude Skills Implements 3-Layer Context Architecture to Manage Hundreds of Skills
Anthropic's Claude Skills framework employs a three-layer context management system that loads only skill metadata by default, enabling support for hundreds of specialized skills without exceeding context window limits.
How Weaviate Agent Skills Let Claude Code Build Vector Apps in Minutes
Weaviate's official Agent Skills give Claude Code structured access to vector databases, eliminating guesswork when building semantic search and RAG applications.
Boris Cherny's Claude Code Tips Are Now a Skill. Here Is What the Complete Collection Reveals.
A curated collection of expert Claude Code tips is now available as a shareable 'Skill,' revealing proven workflows for faster, more reliable agentic coding.
HuggingFace Launches Daily Papers SKILL.md for AI Agents to Read, Search, and Fetch Research Papers
HuggingFace released Daily Papers SKILL.md, a tool enabling AI agents to read paper content as markdown, search papers, find linked models/datasets, and fetch papers via API.
How Adding 'Skills' to MCP Tools Cuts Agent Token Usage by 87%
Adding structured 'skills' descriptions to MCP tools dramatically reduces token consumption in custom agents—here's how to implement it in your Claude Code workflows.
Build Self-Evolving Skills for Claude Code: The GitHub Pattern That Grows Smarter With Use
A new GitHub pattern shows how to create Claude Code Skills that learn from each session, preventing knowledge loss and reducing repetitive context.
Stanford-Princeton Team Open-Sources LabClaw: The 'Skill OS' for Scientific AI
Researchers from Stanford and Princeton have open-sourced LabClaw, a 'Skill Operating Layer' for LabOS that transforms natural language commands into executable lab workflows. This breakthrough promises to dramatically accelerate scientific experimentation by bridging human intent with robotic execution.
EvoSkill: How AI Agents Are Learning to Teach Themselves New Skills
Researchers have developed EvoSkill, a self-evolving framework where AI agents automatically discover and refine their own capabilities through failure analysis. The system improves performance by up to 12% on complex tasks and demonstrates skill transfer between different domains.
Anthropic Releases Comprehensive Guide to Building Custom AI Skills for Claude
Anthropic has published a detailed 33-page guide for developers to create custom skills for Claude AI. This cheat sheet teaches how to package instructions into folders that enable Claude to handle specific tasks and workflows, representing a major step in AI customization.
Beyond Simple Search: How Advanced Image Retrieval Transforms Luxury Discovery
New research reveals major flaws in current visual search tech. For luxury retail, this means missed sales from poor multi-item inspiration and inconsistent results. A new benchmark and method promise more accurate, nuanced product discovery.
Optimizing Luxury Discovery: A Smarter Pre-Ranking Engine for Personalization
New research tackles inefficiency in recommendation pipelines by intelligently separating 'easy' from 'hard' customer matches. This heterogeneity-aware pre-ranking can boost personalization accuracy while controlling computational costs, directly applicable to luxury product discovery and clienteling.
SkillsMP Launches AI 'App Store' with 270,000+ Claude Skills for Seamless Code Automation
SkillsMP introduces an open-source marketplace with over 270,000 specialized AI skills for Claude Code, enabling automatic skill invocation without manual prompting. The platform eliminates setup friction while supporting cross-model compatibility through an open standard.
GitHub Repository Unleashes 1,715+ Production-Ready AI Agent Skills
A new GitHub repository has surfaced containing over 1,715 production-ready AI agent skills that developers can install and deploy in seconds. This collection represents a significant leap in accessible AI tooling, potentially accelerating agent-based application development across industries.
Add 197 Bioinformatics Skills to Claude Code with SciAgent-Skills
A ready-to-use plugin that transforms Claude Code into a bioinformatics expert without fine-tuning or RAG setup.
AI's 'Hollowing Out' Effect: How Automation Targets High-Value, High-Skill Tasks First
A viral commentary by George Pu posits that AI's primary impact isn't mass job elimination but the systematic automation of a role's most valuable, specialized, and well-compensated tasks, leaving workers with diminished, less critical duties.
Beyond Basic Browsing: Adaptive Multimodal AI for Next-Gen Luxury Discovery
A new AI model, CAMMSR, dynamically fuses image, text, and sequence data to understand nuanced client preferences. For luxury retail, this enables hyper-personalized recommendations that adapt to a client's evolving taste across categories, boosting engagement and conversion.
Omar Sarayra Builds LLM Artifact Generator for AI Knowledge Discovery
Omar Sarayra created a system that transforms dense LLM knowledge bases into consumable visual artifacts, like a pulse on HN AI discussions. He argues this format could become a new medium for staying current.
PRL-Bench: LLMs Score Below 50% on End-to-End Physics Research Tasks
Researchers introduced PRL-Bench, a benchmark built from 100 recent Physical Review Letters papers, testing LLMs on end-to-end physics research. Top models scored below 50%, exposing a significant capability gap for autonomous scientific discovery.
Akshay Pachaar Inverts LLM Agent Architecture with 'Harness' Design
AI engineer Akshay Pachaar outlined a novel 'harness' architecture for LLM agents that externalizes intelligence into memory, skills, and protocols. He is building a minimal, didactic open-source implementation of this design.
AI-Powered 'Vibe Coding' Drives 84% Surge in App Store Submissions
App Store submissions surged 84% last year to over 600,000 new apps, driven by AI-assisted 'vibe coding.' This rapid proliferation is devaluing traditional development skills and flooding the market with low-quality applications.
Anthropic Launches Project Glasswing for Critical Software Security
Anthropic announced Project Glasswing, an urgent initiative to secure critical software, powered by its new frontier model Claude Mythos Preview, which it claims can find vulnerabilities better than all but the most skilled humans.
AWP (Agent Work Protocol) Launches Testnet on Base, Enabling Autonomous AI Agent Work Coordination
Developer hasantoxr has launched AWP, an open protocol on Base testnet that allows AI agents to autonomously register, find work, and execute tasks without human prompting. The system uses skill files to define work types, enabling gasless agent coordination.
Learning to Disprove: LLMs Fine-Tuned for Formal Counterexample Generation in Lean 4
Researchers propose a method to train LLMs for formal counterexample generation, a neglected skill in mathematical AI. Their symbolic mutation strategy and multi-reward framework improve performance on three new benchmarks.
Building a Smart Learning Path Recommendation System Using Graph Neural Networks
A technical article outlines how to build a learning path recommendation system using Graph Neural Networks (GNNs). It details constructing a knowledge graph and applying GNNs for personalized course sequencing, a method with clear parallels to retail product discovery.
Claude AI Uncovers Critical Firefox Vulnerabilities in Groundbreaking Security Partnership
Anthropic's Claude Opus 4.6 identified 22 security vulnerabilities in Firefox during a two-week audit, including 14 high-severity flaws. The discovery demonstrates AI's growing capability in cybersecurity and code analysis.
Beyond Cosine Similarity: How Embedding Magnitude Optimization Can Transform Luxury Search & Recommendation
New research reveals that controlling embedding magnitude—not just direction—significantly boosts retrieval and RAG performance. For luxury retail, this means more accurate product discovery, personalized recommendations, and enhanced clienteling through superior semantic search.