Agent-Agnostic
    skill.md
    ai agents
    compatibility

    Every AI Coding Agent That Supports SKILL.md (2026)

    Complete list of AI coding agents that support SKILL.md in 2026. Claude Code, OpenClaw, Codex CLI, Cursor, Gemini CLI, and more.

    April 23, 20265 min read
    Share:

    Every AI Coding Agent That Supports SKILL.md (2026)

    SKILL.md is an open file format that teaches AI coding agents new capabilities. It started as Anthropic's format for Claude Code and has been adopted across a growing ecosystem. Here's every agent that supports it.

    Quick Answer: As of 2026, SKILL.md is supported by Claude Code, OpenClaw, Codex CLI, Cursor, Gemini CLI, and 15+ other AI coding agents. Skills are portable — the same SKILL.md file works across all compatible agents without modification. Browse ready-to-install skills at agensi.io/skills.

    Which agents support SKILL.md?

    AgentDeveloperSKILL.md SupportPersonal Skills PathProject Skills Path
    Claude CodeAnthropicFull~/.claude/skills/.claude/skills/
    OpenClawPeter SteinbergerFull~/.openclaw/skills/.openclaw/skills/
    Codex CLIOpenAIFull (+ openai.yaml)~/.codex/skills/.codex/skills/
    CursorCursor IncPartial (project-only)N/A.cursor/skills/
    Gemini CLIGoogleFull~/.gemini/skills/.gemini/skills/

    Beyond these major agents, the SKILL.md format has been adopted by a growing number of independent and open-source agents. The format's simplicity — a markdown file with YAML frontmatter — makes it easy for any agent to implement.

    How does cross-agent compatibility work?

    SKILL.md is a standardized format. The core elements are universal:

    YAML frontmatter with name and description fields. Every agent reads these to identify the skill and decide when to activate it.

    Markdown body with instructions. Every agent reads this as the skill's operating instructions.

    Folder structure. Every agent expects a folder containing a SKILL.md file, optionally with supporting files (scripts, references).

    This means a skill built for Claude Code works in OpenClaw, Codex CLI, and other agents. You copy the folder to the correct directory and it works.

    What are the limitations of cross-agent skills?

    While the core format is portable, some agent-specific features don't translate:

    Claude Code's context: fork runs skills as subagents with isolated context. OpenClaw and other agents may handle this differently or ignore it.

    Codex CLI's openai.yaml adds UI metadata and MCP tool dependencies. Other agents ignore this file entirely.

    Cursor's project-only scope means personal skills from other agents need to be copied into each project.

    For most skills — code review, testing, git automation, documentation — these differences don't matter. The instructions work the same across agents.

    One subscription for every agent.

    Agensi Pro provides MCP access to the full skill catalog. Works with Claude Code, OpenClaw, Codex CLI, and any agent that supports MCP.

    Start 3-day free trial →

    $9/mo after trial. Card required. Cancel anytime.

    Why does cross-agent support matter?

    The AI coding agent market is evolving quickly. New agents launch regularly, existing ones add features, and developers switch between tools. SKILL.md provides insurance: your investment in skills isn't locked to one vendor.

    If you build a custom code review skill for your team using Claude Code and later switch to OpenClaw or Codex CLI, the skill moves with you. Zero migration effort.

    This portability is also why Agensi exists as a cross-agent marketplace. Skills aren't sold for a specific agent — they work across the entire ecosystem.

    How do I install skills for different agents?

    Each agent reads from its own directory. Install the same skill across multiple agents by copying:

    # Download once from Agensi
    # Then install in each agent's directory:
    
    cp -r code-reviewer ~/.claude/skills/
    cp -r code-reviewer ~/.openclaw/skills/
    cp -r code-reviewer ~/.codex/skills/
    cp -r code-reviewer .cursor/skills/
    cp -r code-reviewer ~/.gemini/skills/
    

    Or use Agensi Pro for MCP-based access, which works with any agent that supports the Model Context Protocol.

    What is the Agent Skills open standard?

    SKILL.md is part of a broader specification called the Agent Skills open standard. The specification defines the file format, frontmatter fields, directory conventions, and discovery mechanisms that agents implement.

    The standard was originally created by Anthropic for Claude Code and published openly so other agents could adopt it. The goal is interoperability — a single skill format that works everywhere, preventing vendor lock-in in the AI coding tools market.

    For the full format specification, read SKILL.md Format Reference. For an introduction to the standard, read What Is SKILL.md?.


    Browse security-scanned skills for every SKILL.md-compatible agent on Agensi.

    Find the right skill for your workflow

    Browse our marketplace of AI agent skills, ready to install in seconds.

    Browse Skills

    Related Articles