agentready.json v0.1 is the stable output surface for CI and review.

Required root fields

v0.1 requires agentready_version, source_type, source_name, generated_at, score, status, risk_counts and tools. Current exports may include compatibility aliases such as source and summary.

Tool fields

Each tool or operation exposes id, name, method, path, action_type, severity, detected_risks, rule_codes, detected_rules and recommendations.

Minimal complete example

{
  "agentready_version": "0.1",
  "source_type": "openapi",
  "source_name": "billing-openapi.json",
  "generated_at": "2026-07-10T05:30:00.000Z",
  "score": 90,
  "status": "AgentReady",
  "risk_counts": {
    "critical": 0,
    "high": 0,
    "medium": 0,
    "low": 0
  },
  "tools": []
}

PASS and FAIL policy

PASS means the score is at least min-score and no finding at or above fail-on exists. FAIL means the score is below the policy threshold or at least one finding meets the severity gate.

• Exit code 0: policy PASS.
• Exit code 1: policy FAIL.
• Exit code 2: invalid input or CLI usage error.
• Exit code 3: unexpected internal error.

Score and status

The V1 score is 0 to 100. Status bands are 85-100 AgentReady, 70-84 Minor fixes, 50-69 Needs fixes and 0-49 Not AgentReady. CI blocking can still fail a high score when fail-on detects a severe finding.

Versioning and freshness

Results should be read together with engine version, ruleset version, AgentReady schema version, policy version, source protocol or version, input hash, commit and scan date when available. Historical results can become stale when rules, policies or source contracts change.