breakpilot-compliance/backend-compliance/compliance/api/agent_check/_b17_wiring.py

"""B17 wiring — Audit-Walk-Recorder.

Triggert beim consent-tester einen kompletten Playwright-Site-Walk
mit Video-Aufzeichnung. Result: Video + JSON-Action-Index mit
Timestamps + SHA-256-Hash für Manipulation-Schutz.

Speichert nur die Walk-Metadata + Video-URL im state. Der eigentliche
File-Body bleibt im consent-tester-Volume (Stufe 1). Stufe 3 wird das
Video zu DSMS-IPFS hochladen und die CID hier einbinden.
"""

from __future__ import annotations

import html
import logging
from urllib.parse import urlparse

import httpx

from ._constants import CONSENT_TESTER_URL

logger = logging.getLogger(__name__)


async def run_b17(state: dict) -> None:
    """Trigger walk recording + store metadata in state."""
    req = state.get("req")
    if req is None:
        return
    homepage = ""
    for d in req.documents:
        if d.url:
            p = urlparse(d.url)
            if p.scheme and p.netloc:
                homepage = f"{p.scheme}://{p.netloc}/"
                break
    if not homepage:
        return

    walk: dict = {}
    try:
        async with httpx.AsyncClient(timeout=180.0) as c:
            r = await c.post(
                f"{CONSENT_TESTER_URL}/scan-audit-walk",
                json={"url": homepage, "dwell_s": 4.0, "max_links": 8},
                timeout=180.0,
            )
            if r.status_code == 200:
                walk = r.json()
    except Exception as e:
        logger.warning("B17 audit-walk request failed: %s", e)
        return

    if not walk or not walk.get("walk_id"):
        return

    state["audit_walk"] = walk
    state["audit_walk_html"] = _render(walk)
    logger.info(
        "B17 audit-walk: %s · %d actions · video %d bytes · sha256 %s",
        walk.get("walk_id"),
        len(walk.get("actions") or []),
        (walk.get("video") or {}).get("size_bytes", 0),
        ((walk.get("video") or {}).get("sha256") or "")[:12],
    )


def _video_link(walk_id: str) -> str:
    """External URL for the recorded video (when consent-tester is
    reachable from the audit reviewer)."""
    return f"{CONSENT_TESTER_URL}/audit-walks/{walk_id}/video.webm"


def _render(walk: dict) -> str:
    wid = walk.get("walk_id") or ""
    video = walk.get("video") or {}
    actions = walk.get("actions") or []
    nav_count = sum(1 for a in actions if a.get("action") == "navigate")
    sha = (video.get("sha256") or "")[:12]
    size_kb = round((video.get("size_bytes") or 0) / 1024, 1)
    walk_link = _video_link(wid)
    meta_link = f"{CONSENT_TESTER_URL}/audit-walks/{wid}/walk.json"

    rows = []
    for a in actions:
        ts = (a.get("timestamp") or "")[11:19]  # HH:MM:SS
        act = a.get("action") or ""
        detail = ""
        if act == "goto" or act == "navigate":
            detail = (a.get("url") or "")[:120]
            if a.get("status"):
                detail += f" → HTTP {a['status']}"
        elif act == "accept_banner":
            r = a.get("result") or ""
            if r == "clicked":
                detail = f"Banner akzeptiert ({a.get('phrase') or a.get('selector') or ''})"
            else:
                detail = "Kein Accept-Button gefunden"
        elif act == "discover_footer_links":
            detail = f"{a.get('count', 0)} Compliance-Links im Footer"
        rows.append(
            f"<tr><td style='padding:4px 8px;font-family:monospace;"
            f"color:#475569;'>{html.escape(ts)}</td>"
            f"<td style='padding:4px 8px;'>{html.escape(act)}</td>"
            f"<td style='padding:4px 8px;color:#475569;'>"
            f"{html.escape(detail)}</td></tr>"
        )
    return (
        "<div style='margin:24px 0;padding:16px;border-left:4px solid #0ea5e9;"
        "background:#f0f9ff;border-radius:4px;'>"
        "<h2 style='margin:0 0 8px;color:#0c4a6e;font-size:16px;'>"
        "🎥 Audit-Walk-Video (Beweis-Aufzeichnung)"
        "</h2>"
        "<p style='margin:0 0 8px;font-size:13px;color:#475569;'>"
        f"<strong>Video:</strong> "
        f"<a href='{html.escape(walk_link)}' style='color:#0369a1;'>video.webm</a> "
        f"({size_kb} KB, SHA-256 <code>{html.escape(sha)}…</code>) · "
        f"<strong>Metadata:</strong> "
        f"<a href='{html.escape(meta_link)}' style='color:#0369a1;'>walk.json</a>"
        "</p>"
        "<p style='margin:0 0 8px;font-size:13px;color:#475569;'>"
        f"{nav_count} Compliance-Seiten besucht, jede 4 Sek "
        "verweilt — Reviewer kann den Audit-Walk nachverfolgen."
        "</p>"
        "<table style='font-size:12px;width:100%;border-collapse:collapse;"
        "background:#fff;border-radius:4px;'>"
        "<thead><tr style='background:#e0f2fe;'>"
        "<th style='padding:6px 8px;text-align:left;'>Zeit (UTC)</th>"
        "<th style='padding:6px 8px;text-align:left;'>Aktion</th>"
        "<th style='padding:6px 8px;text-align:left;'>Detail</th>"
        "</tr></thead><tbody>" + "".join(rows) + "</tbody></table>"
        "</div>"
    )