CLI contract

Resolve, audit, and install agent skills from the terminal.

OpenAgentSkill CLI turns a natural-language task into a selected skill, safety policy decision, target-specific install handoff, and audit trail.

View API contract Try resolve endpoint

Resolve

Resolve a task into one safe install plan

Calls /api/agent/resolve and returns the selected skill, alternatives, safety profile, and install prompt.

openagentskill resolve "scrape competitor pricing pages" --agent codex --max-risk medium

Install

Fetch target-specific install handoffs

Returns Codex, Claude Code, Cursor, and CLI install handoffs from the canonical skill record.

openagentskill install crawl4ai --agent claude-code

Evaluate

Check registry recommendation quality

Runs the public recommendation benchmark so ranking changes can be checked before deployment.

openagentskill evals

Resolve API

The CLI is a thin layer over the agent control plane.

The important product surface is the decision object: selected skill, alternatives, policy result, safety score, install plan, and benchmark link. Agents can call the same endpoint directly without opening a browser.

Request

{
  "task": "review a pull request and summarize risky changes",
  "agent": "codex",
  "constraints": {
    "max_risk": "medium",
    "needs_install_command": true,
    "min_stars": 500
  }
}

Future lockfile

{
  "version": 1,
  "skills": {
    "crawl4ai": {
      "source": "github.com/unclecode/crawl4ai",
      "audit_score": 92,
      "safety_score": 80,
      "resolved_at": "2026-06-10T00:00:00.000Z"
    }
  }
}