{
  "schema_version": "1.1",
  "id": "archive:https://www.youtube.com/watch?v=26RtyAm9y_Q",
  "slug": "the-dark-arts-of-web-automation-teaching-agents-to-use-websites-like-hum-12nj0e0",
  "url": "https://feed7.dev/p/the-dark-arts-of-web-automation-teaching-agents-to-use-websites-like-hum-12nj0e0",
  "title": "The Dark Arts of Web Automation: Teaching Agents to Use Websites Like Humans — Corey Gallon, Rexmore",
  "why_included": "Reliable browser agents pair a programmable CLI with Chrome DevTools Protocol input, then sense, act, and verify one move at a time. Save solved paths as deterministic code instead of re-reasoning.",
  "summary": "The approach gives agents a CLI over Chrome DevTools Protocol, exposing the DOM, accessibility tree, screenshots, network traffic, logs, clicks, and keystrokes. Its core control loop is **sense, act, verify** across independent channels.",
  "practical_implication": "Start with cheap programmatic actions, escalate to trusted browser input only when required, and use vision or human-like motion as the last rung. Once a path works, encode it so deterministic code handles repeated interaction while the model handles perception.",
  "agent_context": "The approach gives agents a CLI over Chrome DevTools Protocol, exposing the DOM, accessibility tree, screenshots, network traffic, logs, clicks, and keystrokes. Its core control loop is **sense, act, verify** across independent channels.\n\nStart with cheap programmatic actions, escalate to trusted browser input only when required, and use vision or human-like motion as the last rung. Once a path works, encode it so deterministic code handles repeated interaction while the model handles perception.\n\nThe demonstrations include bypassing **Turnstile, MTCaptcha, Lemon, and reCAPTCHA v2**, behavior that can trigger abuse controls and violate site policies. The method therefore needs explicit authorization and tight operational boundaries.",
  "source": {
    "name": "AI Engineer",
    "url": "https://www.youtube.com/watch?v=26RtyAm9y_Q",
    "published_at": "2026-08-14T15:30:00.000Z"
  },
  "source_class": "video",
  "content_type": "Video",
  "layer": "agent",
  "domains": [
    "coding"
  ],
  "topics": [
    "computer-use",
    "harness-engineering",
    "tool-use"
  ],
  "verification": {
    "status": "source_linked",
    "label": "Source Linked",
    "method": "source_feed",
    "verified_at": null
  },
  "uncertainty": [
    "The demonstrations include bypassing **Turnstile, MTCaptcha, Lemon, and reCAPTCHA v2**, behavior that can trigger abuse controls and violate site policies. The method therefore needs explicit authorization and tight operational boundaries."
  ],
  "connected_context": {
    "meaning": "This makes browser automation a layered harness problem: combine independent observations, verify each action, and convert successful paths into deterministic routines. It complements visual-only feedback with DOM, accessibility, network, and log evidence, while its CAPTCHA demonstrations sharply increase the need for authorization, action boundaries, and approval controls.",
    "corpus_size": 479,
    "generated_at": "2026-08-18T10:04:58.271Z",
    "connections": [
      {
        "title": "Perception Agents — Antje Barth, Amazon AGI Lab",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=2JX6JYyQG4Y",
        "feed7_url": "https://feed7.dev/p/perception-agents-antje-barth-amazon-agi-lab-1gq6jg4",
        "reason": "The perception harness supplies one feedback channel; this Signal broadens verification across visual, DOM, accessibility, network, and log evidence."
      },
      {
        "title": "From RL to IRL — Gaurav Mishra, Amazon AGI Lab",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=Cc0_nyxROBA",
        "feed7_url": "https://feed7.dev/p/from-rl-to-irl-gaurav-mishra-amazon-agi-lab-0ex196x",
        "reason": "Its sense-act-verify loop addresses ordinary interaction failures, while the candidate adds checkpoints, rollback, interruption, and handoff for persistent or risky consequences."
      },
      {
        "title": "Bringing agents onto the world wide web — Paul Klein IV, Browserbase",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=GqoNrUz8hEU",
        "feed7_url": "https://feed7.dev/p/bringing-agents-onto-the-world-wide-web-paul-klein-iv-browserbase-1vea3od",
        "reason": "Both place browser reliability in the runtime harness; the candidate adds stable environments, authentication, site knowledge, and approvals needed around this interaction ladder."
      },
      {
        "title": "In Code They Act, In Proof We Trust — Erik Meijer, Leibniz Labs",
        "source_name": "YouTube",
        "source_url": "https://www.youtube.com/watch?v=-CnA2lGfymY",
        "feed7_url": "https://feed7.dev/p/in-code-they-act-in-proof-we-trust-erik-meijer-leibniz-labs-1lyno2y",
        "reason": "Encoding proven paths as deterministic code makes actions more inspectable, but the candidate imposes the stronger prerequisite of machine-checkable safety before side effects execute."
      }
    ]
  },
  "lifecycle": "Current",
  "published_at": "2026-08-14T15:30:00.000Z",
  "modified_at": "2026-08-14T15:30:00.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/the-dark-arts-of-web-automation-teaching-agents-to-use-websites-like-hum-12nj0e0",
    "json": "https://feed7.dev/p/the-dark-arts-of-web-automation-teaching-agents-to-use-websites-like-hum-12nj0e0.json",
    "markdown": "https://feed7.dev/p/the-dark-arts-of-web-automation-teaching-agents-to-use-websites-like-hum-12nj0e0.md"
  }
}