# Stop Your Agent From… — eleven agent incident classes and the fix for each | Series

rm -rf, deleted production databases, leaked .env secrets, runaway API bills, unapproved payments, IAM escalation — the incidents production agents cause, and the three steps that stop each one.

<div class="acp-section" style="padding-top:56px;">
  <div class="acp-container" style="text-align:center;">
    <div style="display:inline-block;padding:5px 16px;border-radius:100px;background:linear-gradient(135deg,rgba(79,70,229,0.14),rgba(6,182,212,0.10));color:var(--color-accent);font-size:12px;font-weight:700;letter-spacing:0.06em;text-transform:uppercase;margin-bottom:16px;">Incident series</div>
    <h1 class="acp-hero-h1" style="font-size:44px;margin:0 auto;max-width:780px;">Stop Your Agent From&hellip;</h1>
    <p class="acp-subtitle" style="max-width:680px;margin:18px auto 0;">
      Every post in this series starts from an incident that actually happened to someone — a wiped home directory, a dropped production database, secrets in a commit, an $18,000 surprise bill — and ends with the exact configuration that stops it. Three steps each, a few minutes of work, no smarter prompting required.
    </p>
    <p class="acp-subtitle" style="max-width:680px;margin:14px auto 0;font-size:14px;">
      Wondering what your agent could do beyond these? The <a href="/tool-surfaces">Tool Surface Index</a> lists every tool Claude Code and Codex declare, grouped by blast radius.
    </p>
  </div>
</div>

<div class="acp-section" style="padding-top:8px;">
  <div class="acp-container" style="max-width:780px;">
    <div class="acp-card acp-card-pad">
      <h2 class="acp-section-title" style="margin-bottom:8px;">Posts in this series</h2>
      <p class="acp-muted" style="margin-bottom:24px;font-size:14px;">Each post covers one incident class: what happened, why it's one tool call away in your setup, and the three steps that close it.</p>
      <ol style="list-style:none;counter-reset:series-counter;padding:0;margin:0;display:grid;gap:18px;">
        
        
        <li style="counter-increment:series-counter;display:grid;grid-template-columns:32px 1fr;gap:16px;padding-bottom:18px;border-bottom:1px solid var(--color-border);">
          <span style="font-family:var(--font-mono);font-size:13px;color:var(--color-accent);font-weight:700;padding-top:2px;">.</span>
          <div>
            <a href="" style="font-size:17px;font-weight:600;color:var(--color-text-primary);text-decoration:none;line-height:1.35;"></a>
            <div class="acp-muted" style="font-size:14px;margin-top:6px;line-height:1.5;"></div>
          </div>
        </li>
        
      </ol>
      <p class="acp-muted" style="margin-top:24px;font-size:14px;">
        All eleven fixes ride the same install — see <a href="/getting-started">getting started</a>, or the one-command setup for <a href="/for-coding-agents">coding agents specifically</a>.
      </p>
    </div>
  </div>
</div>

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "CollectionPage",
  "name": "Stop Your Agent From… — eleven agent incident classes and the fix for each",
  "url": "https://agenticcontrolplane.com/series/stop-your-agent",
  "description": "Eleven real AI-agent incident classes — destructive shell commands, deleted databases, leaked secrets, runaway API spend, unapproved payments, IAM escalation — each with the three-step configuration that prevents it."
}
</script>
