Agent risk · law 4

Stripe

agent: at_risk score 20 posture: stop

Do not rely on this tool right now.

Open risks

the output is no longer provably the declared shape — your agent may act on unverified data — tool degraded: POST /v1/terminal/readers/{reader}/cancel_action(open 12d ago)

For your agent — poll this

https://api.verifyport.dev/v1/apis/stripe/agent-risk

{
  "slug": "stripe",
  "agent_trust": "at_risk",
  "agent_trust_score": 20,
  "posture": "stop",
  "open_risks": [
    {
      "tool": "POST /v1/terminal/readers/{reader}/cancel_action",
      "agent_impact": "unverified_output",
      "since": "2026-07-30T02:28:16Z",
      "summary": "tool degraded: POST /v1/terminal/readers/{reader}/cancel_action"
    }
  ]
}

A tool-router can gate on posture (trust · caution · pin-version · stop) in one fetch — exactly what this panel shows.