{
  "schema_version": "newruntime-topic-hub-v0.1",
  "type": "topic_hub",
  "slug": "skills",
  "title": "Agent skills - New Runtime",
  "description": "Portable skill files, behavioral contracts, tool instructions, and evals that make agent capability reusable.",
  "answer": [
    "Agent skills package repeatable behavior outside the current chat.",
    "The important boundary is whether the skill changes task outcomes, not whether the instruction prose looks good.",
    "Skills become infrastructure when they are versioned, evaluated, and shared across agents."
  ],
  "search_intents": [
    "agent skills",
    "AI agent skill files",
    "reusable agent capabilities"
  ],
  "status": "featured",
  "last_updated": "2026-07-04",
  "counts": {
    "total": 13,
    "signals": 13,
    "patterns": 0,
    "posts": 0,
    "atlas": 0,
    "sources": 22
  },
  "routes": {
    "html": "https://newruntime.com/topics/skills/",
    "markdown": "https://newruntime.com/topics/skills.md",
    "json": "https://newruntime.com/topics/skills.json"
  },
  "top_sources": [
    "https://claude.com/blog/steering-claude-code-skills-hooks-rules-subagents-and-more",
    "https://cursor.com/marketplace/cursor/cursor-team-kit",
    "https://developers.openai.com/codex/record-and-replay",
    "https://github.com/anthropics/claude-for-legal",
    "https://github.com/emilkowalski/skills",
    "https://github.com/mvanhorn/last30days-skil",
    "https://github.com/mvanhorn/last30days-skill.git",
    "https://github.com/NousResearch/hermes-agent/blob/main/optional-skills/creative/creative-ideation/SKILL.md",
    "https://github.com/nvidia/skillspector",
    "https://github.com/openai/skills/tree/main/skills/.curated/gh-address-comments",
    "https://github.com/senlindesign/taste-skill",
    "https://huggingface.co/"
  ],
  "patterns": [],
  "field_notes": [],
  "raw_signals": [
    {
      "kind": "Raw Signal",
      "slug": "animation-vocabulary-turns-motion-references-into-an-agent-skill",
      "title": "Animation Vocabulary Turns Motion References into an Agent Skill",
      "description": "A compact design skill maps visual intent to named motion patterns, giving interface agents a reusable language for implementing animation decisions.",
      "date": "2026-07-04",
      "topics": [
        "design-systems",
        "animation",
        "skills"
      ],
      "source_count": 1,
      "metric": "notable",
      "url": "https://newruntime.com/signals/animation-vocabulary-turns-motion-references-into-an-agent-skill/"
    },
    {
      "kind": "Raw Signal",
      "slug": "a-github-skill-can-resolve-review-comments-as-a-batch",
      "title": "A GitHub Skill Can Resolve Review Comments as a Batch",
      "description": "A reusable Codex workflow gathers pull-request feedback, plans changes, applies fixes, and reports resolution state as one auditable review task.",
      "date": "2026-07-03",
      "topics": [
        "code-review",
        "skills",
        "github"
      ],
      "source_count": 1,
      "metric": "notable",
      "url": "https://newruntime.com/signals/a-github-skill-can-resolve-review-comments-as-a-batch/"
    },
    {
      "kind": "Raw Signal",
      "slug": "creative-ideation-becomes-a-routable-agent-skill",
      "title": "Creative Ideation Becomes a Routable Agent Skill",
      "description": "A packaged workflow decomposes creative exploration into reference gathering, divergence, critique, and selection instead of asking a model to be creative in one step.",
      "date": "2026-06-27",
      "topics": [
        "creative-work",
        "skills",
        "design"
      ],
      "source_count": 1,
      "metric": "notable",
      "url": "https://newruntime.com/signals/creative-ideation-becomes-a-routable-agent-skill/"
    },
    {
      "kind": "Raw Signal",
      "slug": "agentic-resource-discovery-indexes-skills-mcp-and-agents",
      "title": "Agentic Resource Discovery Indexes Skills, MCP, and Agents",
      "description": "Hugging Face treats agent capabilities as searchable resources, creating a discovery layer for tools that can be loaded at runtime.",
      "date": "2026-06-24",
      "topics": [
        "agent-discovery",
        "mcp",
        "skills"
      ],
      "source_count": 3,
      "metric": "notable",
      "url": "https://newruntime.com/signals/agentic-resource-discovery-indexes-skills-mcp-and-agents/"
    },
    {
      "kind": "Raw Signal",
      "slug": "hermes-turns-solved-tasks-into-new-skills",
      "title": "Hermes Turns Solved Tasks into New Skills",
      "description": "An agent runtime can preserve successful procedures as reusable capability modules, converting execution traces into a compounding operational library.",
      "date": "2026-06-24",
      "topics": [
        "hermes",
        "skills",
        "agent-learning"
      ],
      "source_count": 1,
      "metric": "notable",
      "url": "https://newruntime.com/signals/hermes-turns-solved-tasks-into-new-skills/"
    },
    {
      "kind": "Raw Signal",
      "slug": "claude-md-shrinks-as-skills-hooks-and-subagents-mature",
      "title": "CLAUDE.md Shrinks as Skills, Hooks, and Subagents Mature",
      "description": "Agent behavior moves out of one giant context file into scoped mechanisms that load only for the relevant task and lifecycle stage.",
      "date": "2026-06-23",
      "topics": [
        "claude-code",
        "skills",
        "context-engineering"
      ],
      "source_count": 1,
      "metric": "notable",
      "url": "https://newruntime.com/signals/claude-md-shrinks-as-skills-hooks-and-subagents-mature/"
    },
    {
      "kind": "Raw Signal",
      "slug": "codex-turns-a-recorded-mac-workflow-into-a-reusable-skill",
      "title": "Codex Turns a Recorded Mac Workflow into a Reusable Skill",
      "description": "Record and Replay captures a human procedure, converts it into agent instructions, and preserves the successful interaction as reusable behavior.",
      "date": "2026-06-22",
      "topics": [
        "codex",
        "skills",
        "computer-use"
      ],
      "source_count": 2,
      "metric": "notable",
      "url": "https://newruntime.com/signals/codex-turns-a-recorded-mac-workflow-into-a-reusable-skill/"
    },
    {
      "kind": "Raw Signal",
      "slug": "taste-lab-uses-a-design-system-as-a-reasoning-brief",
      "title": "Taste Lab Uses a Design System as a Reasoning Brief",
      "description": "Taste Lab packages aesthetic decisions and implementation rules as agent-readable context rather than relying on vague style prompts.",
      "date": "2026-06-19",
      "topics": [
        "design-systems",
        "skills",
        "frontend"
      ],
      "source_count": 3,
      "metric": "notable",
      "url": "https://newruntime.com/signals/taste-lab-uses-a-design-system-as-a-reasoning-brief/"
    },
    {
      "kind": "Raw Signal",
      "slug": "last30days-packages-social-research-as-an-agent-skill",
      "title": "Last30Days Packages Social Research as an Agent Skill",
      "description": "A reusable research workflow gathers recent public discussion, filters repetition, and produces a time-bounded evidence set for another agent or analyst.",
      "date": "2026-06-18",
      "topics": [
        "research-agents",
        "skills",
        "social-data"
      ],
      "source_count": 2,
      "metric": "notable",
      "url": "https://newruntime.com/signals/last30days-packages-social-research-as-an-agent-skill/"
    },
    {
      "kind": "Raw Signal",
      "slug": "skillspector-adds-a-security-gate-for-agent-skills",
      "title": "SkillSpector Adds a Security Gate for Agent Skills",
      "description": "NVIDIA's scanner treats installable agent instructions as executable supply-chain artifacts that require inspection before use.",
      "date": "2026-06-02",
      "topics": [
        "agent-security",
        "skills",
        "supply-chain"
      ],
      "source_count": 1,
      "metric": "notable",
      "url": "https://newruntime.com/signals/skillspector-adds-a-security-gate-for-agent-skills/"
    },
    {
      "kind": "Raw Signal",
      "slug": "claude-for-legal-packages-a-legal-team-as-plugins",
      "title": "Claude for Legal Packages a Legal Team as Plugins",
      "description": "Anthropic's legal repository expresses domain knowledge, procedures, and tools as installable modules for repeatable professional work.",
      "date": "2026-05-20",
      "topics": [
        "legal-agents",
        "skills",
        "enterprise-ai"
      ],
      "source_count": 1,
      "metric": "notable",
      "url": "https://newruntime.com/signals/claude-for-legal-packages-a-legal-team-as-plugins/"
    },
    {
      "kind": "Raw Signal",
      "slug": "skill-md-acts-as-a-behavior-loader-not-a-better-prompt",
      "title": "SKILL.md Acts as a Behavior Loader, Not a Better Prompt",
      "description": "Skills package procedures, tools, and progressive context so an agent can load behavior only when the task requires it.",
      "date": "2026-05-07",
      "topics": [
        "skills",
        "context-engineering",
        "agent-harness"
      ],
      "source_count": 3,
      "metric": "notable",
      "url": "https://newruntime.com/signals/skill-md-acts-as-a-behavior-loader-not-a-better-prompt/"
    },
    {
      "kind": "Raw Signal",
      "slug": "cursor-ships-its-development-procedures-as-agent-skills",
      "title": "Cursor Ships Its Development Procedures as Agent Skills",
      "description": "Cursor Team Kit packages verification, CI repair, review preparation, and code cleanup as reusable agent behavior.",
      "date": "2026-05-06",
      "topics": [
        "skills",
        "coding-agents",
        "verification"
      ],
      "source_count": 2,
      "metric": "notable",
      "url": "https://newruntime.com/signals/cursor-ships-its-development-procedures-as-agent-skills/"
    }
  ],
  "atlas_records": []
}