luckiest-co 1.0.0 → 1.0.3

package/skills/luckiest-trends/scripts/lib/youtube_yt.py

@@ -0,0 +1,1365 @@
+"""YouTube search and transcript extraction via yt-dlp for the v3.0.0 pipeline.
+
+Uses yt-dlp (https://github.com/yt-dlp/yt-dlp) for both YouTube search and
+transcript extraction. No API keys needed — just have yt-dlp installed.
+
+Inspired by Peter Steinberger's toolchain approach (yt-dlp + summarize CLI).
+"""
+
+import json
+import math
+import os
+import re
+import shlex
+import shutil
+import sys
+import tempfile
+import time
+import urllib.error
+import urllib.request
+from concurrent.futures import ThreadPoolExecutor, as_completed
+from pathlib import Path
+from typing import Any, Dict, List, Optional, Set, Tuple
+
+# Depth configurations: how many videos to search / transcribe
+DEPTH_CONFIG = {
+    "quick": 6,
+    "default": 8,
+    "deep": 40,
+}
+
+TRANSCRIPT_LIMITS = {
+    "quick": 0,
+    "default": 2,
+    "deep": 8,
+}
+
+# Cumulative yt-dlp transcript-fetch stats for the current process. The final
+# report only sees post-pruning items, so it can't distinguish "fetches failed
+# (stale binary)" from "fetches succeeded but the videos were pruned later".
+# quality_nudge reads these via last30days.py to suppress the stale-yt-dlp
+# nudge when every attempted fetch actually succeeded. yt-dlp path only: the
+# nudge diagnoses the local binary, not the ScrapeCreators API.
+_TRANSCRIPT_FETCH_STATS = {"attempts": 0, "failures": 0}
+
+
+def get_transcript_fetch_stats() -> Dict[str, int]:
+    """Return cumulative transcript-fetch stats for this process."""
+    return dict(_TRANSCRIPT_FETCH_STATS)
+
+
+def reset_transcript_fetch_stats() -> None:
+    """Reset cumulative transcript-fetch stats (used by tests)."""
+    _TRANSCRIPT_FETCH_STATS["attempts"] = 0
+    _TRANSCRIPT_FETCH_STATS["failures"] = 0
+
+# Max words to keep from each transcript
+TRANSCRIPT_MAX_WORDS = 5000
+
+from . import dates, http, log, subproc
+from .query import infer_query_intent
+
+from .relevance import token_overlap_relevance as _compute_relevance
+
+# yt-dlp transcript-fetch resilience. A non-zero yt-dlp exit means a real fetch
+# error (rate-limit / bot-check / network), NOT "no captions" — yt-dlp exits 0
+# with no file for a video that genuinely lacks the requested captions. So we
+# capture the returncode, log a classified reason instead of failing silently,
+# and retry transient errors a couple of times with a small per-video staggered
+# backoff.
+_TRANSCRIPT_MAX_RETRIES = 2
+_TRANSCRIPT_BACKOFF_BASE = 2.0  # seconds; multiplied by (attempt + 1)
+_TRANSCRIPT_TIMEOUT = 30  # seconds per yt-dlp attempt (keyless: no fallback to fail over to)
+_TRANSCRIPT_FAST_TIMEOUT = 12  # seconds per attempt when a ScrapeCreators fallback exists
+_SC_LOW_CREDIT_THRESHOLD = 50  # warn once ScrapeCreators credits drop below this
+# Transient = worth retrying (and definitely not "no captions").
+_TRANSIENT_RE = re.compile(
+    r"429|too many requests|sign in to confirm|not a bot|rate.?limit"
+    r"|temporarily|try again|timed out|timeout|connection|unable to (extract|download)"
+    r"|failed to (extract|download)|got error|read error",
+    re.IGNORECASE,
+)
+# A genuine no-captions signal — treat as no captions, never retry/surface.
+_NO_CAPTION_RE = re.compile(
+    r"no subtitles|requested (format|language)|there'?s no .*subtitles",
+    re.IGNORECASE,
+)
+
+
+def extract_transcript_highlights(transcript: str, topic: str, limit: int = 5) -> list[str]:
+    """Extract quotable highlights from a YouTube transcript.
+
+    Filters filler (subscribe, welcome back, etc.), scores sentences by
+    specificity (numbers, proper nouns, topic relevance), and returns
+    the top highlights.
+    """
+    if not transcript:
+        return []
+
+    sentences = re.split(r'(?<=[.!?])\s+', transcript)
+
+    # Fallback for punctuation-free transcripts (common with auto-captions):
+    # chunk into ~20-word segments so they pass the 8-50 word filter.
+    if len(sentences) <= 1 and len(transcript.split()) > 50:
+        words = transcript.split()
+        sentences = [' '.join(words[i:i+20]) for i in range(0, len(words), 20)]
+
+    filler = [
+        r"^(hey |hi |what's up|welcome back|in today's video|don't forget to)",
+        r"(subscribe|like and comment|hit the bell|check out the link|down below)",
+        r"^(so |and |but |okay |alright |um |uh )",
+        r"(thanks for watching|see you (next|in the)|bye)",
+    ]
+
+    topic_words = [w.lower() for w in topic.lower().split() if len(w) > 2]
+
+    candidates = []
+    for sent in sentences:
+        sent = sent.strip()
+        words = sent.split()
+        if len(words) < 8 or len(words) > 50:
+            continue
+        if any(re.search(p, sent, re.IGNORECASE) for p in filler):
+            continue
+
+        score = 0
+        if re.search(r'\d', sent):
+            score += 2
+        if re.search(r'[A-Z][a-z]+', sent):
+            score += 1
+        if '?' in sent:
+            score += 1
+        sent_lower = sent.lower()
+        if any(w in sent_lower for w in topic_words):
+            score += 2
+
+        candidates.append((score, sent))
+
+    candidates.sort(key=lambda x: -x[0])
+    return [sent for _, sent in candidates[:limit]]
+
+
+def _log(msg: str):
+    log.source_log("YouTube", msg, tty_only=False)
+
+
+def is_ytdlp_installed() -> bool:
+    """Check if yt-dlp is available locally, or if SSH routing is configured.
+
+    When LAST30DAYS_YOUTUBE_SSH_HOST is set, returns True without a local check —
+    yt-dlp lives on the remote host. Failures surface naturally on first use.
+    """
+    if _ytdlp_ssh_host():
+        return True
+    return shutil.which("yt-dlp") is not None
+
+
+# Host aliases must be plain hostnames / SSH config aliases — no flags, no
+# shell metacharacters. Rejects any value that could be reinterpreted by ssh
+# (or the surrounding shell) as something other than a destination.
+_SSH_HOST_ALIAS_RE = re.compile(r"^[a-zA-Z0-9._-]+$")
+
+
+def _ytdlp_ssh_host() -> Optional[str]:
+    """Return SSH host alias if yt-dlp should be routed via SSH, else None.
+
+    Set LAST30DAYS_YOUTUBE_SSH_HOST=<ssh-alias> (e.g. 'macmini') in the environment
+    to route yt-dlp through SSH for residential IP egress. This bypasses
+    YouTube's bot-wall on datacenter IPs (Hetzner, DigitalOcean, AWS, etc.)
+    where ytsearch returns 0 results regardless of cookies.
+
+    The remote host must have yt-dlp installed and reachable via the named
+    SSH alias (configured in ~/.ssh/config). On macOS hosts with Homebrew,
+    add brew shellenv to ~/.zshenv (not just ~/.zprofile) so non-login SSH
+    shells find yt-dlp on PATH.
+
+    Validation: host value must match ``[A-Za-z0-9._-]+``. Anything starting
+    with ``-`` or containing shell/SSH metacharacters is rejected with a
+    stderr warning and treated as unset, so a misconfigured or attacker-
+    controlled value can't slip through as an SSH option flag or proxy command.
+    The ``--`` option terminator in ``_wrap_ytdlp_cmd`` is a second line of
+    defense; this regex closes the door on the env var ever reaching ssh
+    in the first place.
+
+    To use a value from ~/.config/last30days/.env, export it into the
+    environment before invoking the engine, e.g. in a wrapper:
+        set -a; source ~/.config/last30days/.env; set +a
+        python3 last30days.py "..."
+    """
+    host = os.environ.get("LAST30DAYS_YOUTUBE_SSH_HOST", "").strip()
+    if not host:
+        return None
+    if not _SSH_HOST_ALIAS_RE.match(host):
+        sys.stderr.write(
+            f"[youtube_yt] WARNING: LAST30DAYS_YOUTUBE_SSH_HOST={host!r} "
+            "does not look like a plain hostname/alias; ignoring. "
+            "Expected pattern: letters, digits, dot, underscore, hyphen.\n"
+        )
+        return None
+    return host
+
+
+def _wrap_ytdlp_cmd(cmd: List[str]) -> List[str]:
+    """Wrap a yt-dlp command list with `ssh <host>` when SSH routing is set.
+
+    Args are shell-quoted to survive the remote shell. Uses BatchMode=yes so
+    a misconfigured key fails fast instead of hanging on a password prompt.
+    The `--` option terminator prevents an SSH option-injection if
+    LAST30DAYS_YOUTUBE_SSH_HOST were ever set to a value starting with `-`.
+    """
+    host = _ytdlp_ssh_host()
+    if not host:
+        return cmd
+    remote_cmd = " ".join(shlex.quote(a) for a in cmd)
+    return ["ssh", "-o", "BatchMode=yes", "--", host, remote_cmd]
+
+
+def _extract_core_subject(topic: str) -> str:
+    """Extract core subject from verbose query for YouTube search.
+
+    NOTE: 'tips', 'tricks', 'tutorial', 'guide', 'review', 'reviews'
+    are intentionally KEPT — they're YouTube content types that improve search.
+    """
+    from .query import VIRAL_NOISE, extract_core_subject
+    # YouTube extends VIRAL_NOISE with temporal/meta words the planner emits
+    # that don't appear in YouTube titles (months, recent year tokens, etc.).
+    _YT_EXTRA = frozenset({
+        'last', 'days', 'recent', 'recently', 'month', 'week',
+        'january', 'february', 'march', 'april', 'may', 'june',
+        'july', 'august', 'september', 'october', 'november', 'december',
+        '2025', '2026', '2027',
+        'music', 'public', 'appearances', 'developments', 'discussions', 'coverage',
+    })
+    return extract_core_subject(topic, noise=VIRAL_NOISE | _YT_EXTRA)
+
+
+def expand_youtube_queries(topic: str, depth: str) -> List[str]:
+    """Generate multiple YouTube search queries from a topic.
+
+    Mirrors reddit.py's expand_reddit_queries() pattern:
+    1. Extract core subject (strip noise words)
+    2. Include original topic if different from core
+    3. Add intent-specific OR-joined content-type variants
+    4. Cap by depth: 1 for quick, 2 for default, 3 for deep
+
+    Returns 1-3 query strings depending on depth.
+    """
+    core = _extract_core_subject(topic)
+    queries = [core]
+
+    # Include cleaned original topic as variant if different from core
+    original_clean = topic.strip().rstrip('?!.')
+    if core.lower() != original_clean.lower() and len(original_clean.split()) <= 8:
+        queries.append(original_clean)
+
+    qtype = infer_query_intent(topic)
+
+    # Intent-specific YouTube content-type variants
+    if qtype == "opinion":
+        queries.append(f"{core} review OR reaction OR breakdown")
+    elif qtype == "product":
+        queries.append(f"{core} review OR comparison OR unboxing")
+    elif qtype == "comparison":
+        queries.append(f"{core} vs OR compared OR head to head")
+    elif qtype == "how_to":
+        queries.append(f"{core} tutorial OR guide OR explained")
+    else:
+        # breaking_news / general — YouTube content types
+        queries.append(f"{core} review OR reaction OR breakdown")
+
+    # Deep depth: add full-length content variant
+    if depth == "deep":
+        queries.append(f"{core} full OR complete OR official")
+
+    # Cap by depth budget
+    caps = {"quick": 1, "default": 2, "deep": 3}
+    cap = caps.get(depth, 2)
+    return queries[:cap]
+
+
+def search_youtube(
+    topic: str,
+    from_date: str,
+    to_date: str,
+    depth: str = "default",
+) -> Dict[str, Any]:
+    """Search YouTube via yt-dlp. No API key needed.
+
+    Args:
+        topic: Search topic
+        from_date: Start date (YYYY-MM-DD)
+        to_date: End date (YYYY-MM-DD)
+        depth: 'quick', 'default', or 'deep'
+
+    Returns:
+        Dict with 'items' list of video metadata dicts.
+    """
+    if not is_ytdlp_installed():
+        return {"items": [], "error": "yt-dlp not installed"}
+
+    count = DEPTH_CONFIG.get(depth, DEPTH_CONFIG["default"])
+    core_topic = _extract_core_subject(topic)
+
+    _log(f"Searching YouTube for '{core_topic}' (since {from_date}, count={count})")
+
+    # yt-dlp search with full metadata (no --flat-playlist so dates are real).
+    # NOTE: --dateafter intentionally omitted — YouTube search returns
+    # relevance-sorted results and strict date filtering returns 0 for
+    # evergreen topics. Python soft filter (below) handles date filtering.
+    cmd = [
+        "yt-dlp",
+        "--ignore-config",
+        "--no-cookies-from-browser",
+        f"ytsearch{count}:{core_topic}",
+        "--dump-json",
+        "--no-warnings",
+        "--no-download",
+    ]
+    cmd = _wrap_ytdlp_cmd(cmd)
+    ssh_host = _ytdlp_ssh_host()
+
+    try:
+        result = subproc.run_with_timeout(cmd, timeout=120)
+    except subproc.SubprocTimeout:
+        _log("YouTube search timed out (120s)")
+        return {"items": [], "error": "Search timed out"}
+    except FileNotFoundError:
+        return {"items": [], "error": "yt-dlp not found"}
+
+    stdout = result.stdout
+    if ssh_host and result.returncode != 0 and not stdout.strip():
+        stderr_first = (result.stderr or "").strip().splitlines()
+        first_line = stderr_first[0] if stderr_first else "(no stderr)"
+        _log(
+            f"YouTube search via SSH host {ssh_host!r} failed "
+            f"(rc={result.returncode}): {first_line}"
+        )
+        return {
+            "items": [],
+            "error": f"SSH routing to {ssh_host!r} failed: {first_line}",
+        }
+    if not stdout.strip():
+        _log("YouTube search returned 0 results")
+        return {"items": []}
+
+    # Parse JSON-per-line output
+    items = []
+    for line in stdout.strip().split("\n"):
+        line = line.strip()
+        if not line:
+            continue
+        try:
+            video = json.loads(line)
+        except json.JSONDecodeError:
+            continue
+
+        video_id = video.get("id", "")
+        view_count = video.get("view_count") if video.get("view_count") is not None else 0
+        like_count = video.get("like_count") if video.get("like_count") is not None else 0
+        comment_count = video.get("comment_count") if video.get("comment_count") is not None else 0
+        upload_date = video.get("upload_date", "")  # YYYYMMDD
+
+        # Convert YYYYMMDD to YYYY-MM-DD
+        date_str = None
+        if upload_date and len(upload_date) == 8:
+            date_str = f"{upload_date[:4]}-{upload_date[4:6]}-{upload_date[6:8]}"
+
+        description = str(video.get("description", ""))[:500]
+        items.append({
+            "video_id": video_id,
+            "title": video.get("title", ""),
+            "url": f"https://www.youtube.com/watch?v={video_id}",
+            "channel_name": video.get("channel", video.get("uploader", "")),
+            "date": date_str,
+            "engagement": {
+                "views": view_count,
+                "likes": like_count,
+                "comments": comment_count,
+            },
+            "duration": video.get("duration"),
+            "relevance": _compute_relevance(core_topic, f"{video.get('title', '')} {description}"),
+            "why_relevant": f"YouTube: {video.get('title', core_topic)[:60]}",
+            "description": description,
+        })
+
+    # Soft date filter: prefer recent items but fall back to all if too few
+    recent = [i for i in items if i["date"] and i["date"] >= from_date]
+    if len(recent) >= 3:
+        items = recent
+        _log(f"Found {len(items)} videos within date range")
+    else:
+        _log(f"Found {len(items)} videos ({len(recent)} within date range, keeping all)")
+
+    # Sort by views descending
+    items.sort(key=lambda x: x["engagement"]["views"], reverse=True)
+
+    return {"items": items}
+
+
+def _clean_vtt(vtt_text: str) -> str:
+    """Convert VTT subtitle format to clean plaintext."""
+    # Strip VTT header
+    text = re.sub(r'^WEBVTT.*?\n\n', '', vtt_text, flags=re.DOTALL)
+    # Strip timestamps
+    text = re.sub(r'\d{2}:\d{2}:\d{2}\.\d{3}\s*-->\s*\d{2}:\d{2}:\d{2}\.\d{3}.*\n', '', text)
+    # Strip position/alignment tags
+    text = re.sub(r'<[^>]+>', '', text)
+    # Strip cue numbers
+    text = re.sub(r'^\d+\s*$', '', text, flags=re.MULTILINE)
+    # Deduplicate overlapping lines
+    lines = text.strip().split('\n')
+    seen = set()
+    unique = []
+    for line in lines:
+        stripped = line.strip()
+        if stripped and stripped not in seen:
+            seen.add(stripped)
+            unique.append(stripped)
+    return re.sub(r'\s+', ' ', ' '.join(unique)).strip()
+
+
+_YT_USER_AGENT = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36"
+
+
+def _fetch_transcript_direct(
+    video_id: str,
+    timeout: int = 30,
+    status: Optional[Dict[str, Any]] = None,
+) -> Optional[str]:
+    """Fetch YouTube transcript via direct HTTP without yt-dlp.
+
+    Scrapes the watch page HTML for the captions track URL in
+    ytInitialPlayerResponse, then fetches the VTT subtitle file.
+
+    Args:
+        video_id: YouTube video ID
+        timeout: HTTP request timeout in seconds
+        status: Optional dict mutated to record per-video signals. Sets
+            ``status["no_caption_tracks"] = True`` when the player response
+            confirms the uploader has no caption tracks (vs. fetch failure).
+
+    Returns:
+        Raw VTT text, or None if captions are unavailable.
+    """
+    watch_url = f"https://www.youtube.com/watch?v={video_id}"
+    headers = {
+        "User-Agent": _YT_USER_AGENT,
+        "Accept-Language": "en-US,en;q=0.9",
+    }
+
+    # Step 1: Fetch the watch page HTML
+    req = urllib.request.Request(watch_url, headers=headers)
+    try:
+        with urllib.request.urlopen(req, timeout=timeout) as resp:
+            html = resp.read().decode("utf-8", errors="replace")
+    except (urllib.error.URLError, urllib.error.HTTPError, OSError, TimeoutError) as exc:
+        _log(f"Direct transcript: failed to fetch watch page for {video_id}: {exc}")
+        return None
+
+    # Step 2: Extract captions URL from ytInitialPlayerResponse
+    # YouTube embeds this as a JS variable in the page HTML
+    match = re.search(
+        r'ytInitialPlayerResponse\s*=\s*(\{.+?\})\s*;(?:\s*var\s|\s*<\/script>)',
+        html,
+    )
+    if not match:
+        # Fallback: try the JSON embedded in the script tag
+        match = re.search(
+            r'var\s+ytInitialPlayerResponse\s*=\s*(\{.+?\})\s*;',
+            html,
+        )
+    if not match:
+        _log(f"Direct transcript: no ytInitialPlayerResponse found for {video_id}")
+        return None
+
+    try:
+        player_response = json.loads(match.group(1))
+    except json.JSONDecodeError:
+        _log(f"Direct transcript: failed to parse ytInitialPlayerResponse for {video_id}")
+        return None
+
+    # Navigate to caption tracks
+    captions = player_response.get("captions", {})
+    renderer = captions.get("playerCaptionsTracklistRenderer", {})
+    caption_tracks = renderer.get("captionTracks", [])
+
+    if not caption_tracks:
+        _log(f"Direct transcript: no caption tracks for {video_id}")
+        if status is not None:
+            status["no_caption_tracks"] = True
+        return None
+
+    # Find English track (prefer exact 'en', then any en variant, then first track)
+    base_url = None
+    for track in caption_tracks:
+        lang = track.get("languageCode", "")
+        if lang == "en":
+            base_url = track.get("baseUrl")
+            break
+    if not base_url:
+        for track in caption_tracks:
+            lang = track.get("languageCode", "")
+            if lang.startswith("en"):
+                base_url = track.get("baseUrl")
+                break
+    if not base_url:
+        # Fall back to first available track
+        base_url = caption_tracks[0].get("baseUrl")
+    if not base_url:
+        _log(f"Direct transcript: no baseUrl in caption tracks for {video_id}")
+        return None
+
+    # Step 3: Fetch the VTT subtitle file
+    sep = "&" if "?" in base_url else "?"
+    vtt_url = f"{base_url}{sep}fmt=vtt"
+    vtt_req = urllib.request.Request(vtt_url, headers=headers)
+    try:
+        with urllib.request.urlopen(vtt_req, timeout=timeout) as resp:
+            vtt_text = resp.read().decode("utf-8", errors="replace")
+    except (urllib.error.URLError, urllib.error.HTTPError, OSError, TimeoutError) as exc:
+        _log(f"Direct transcript: failed to fetch VTT for {video_id}: {exc}")
+        return None
+
+    if not vtt_text or not vtt_text.strip():
+        return None
+
+    return vtt_text
+
+
+def _fetch_transcript_ytdlp_via_ssh(video_id: str, ssh_host: str) -> Optional[str]:
+    """Fetch transcript via yt-dlp on a remote SSH host (mktemp + cat pipeline)."""
+    if not _SSH_HOST_ALIAS_RE.match(ssh_host):
+        return None
+    url = f"https://www.youtube.com/watch?v={video_id}"
+    quoted_url = shlex.quote(url)
+    sub_langs = shlex.quote(_ytdlp_sub_langs())
+    remote_script = (
+        "set -e; "
+        "TMPD=$(mktemp -d); "
+        "yt-dlp --ignore-config --no-cookies-from-browser "
+        f"--write-auto-subs --sub-lang {sub_langs} --sub-format vtt "
+        "--skip-download --no-warnings "
+        f'-o "$TMPD/%(id)s" {quoted_url} >/dev/null 2>&1 || true; '
+        'VTT=$(find "$TMPD" -maxdepth 1 -name "*.vtt" 2>/dev/null | head -1); '
+        '[ -n "$VTT" ] && cat "$VTT"; '
+        'rm -rf "$TMPD"'
+    )
+    cmd = ["ssh", "-o", "BatchMode=yes", "--", ssh_host, remote_script]
+    try:
+        result = subproc.run_with_timeout(cmd, timeout=45)
+    except subproc.SubprocTimeout:
+        _log(f"SSH yt-dlp transcript timed out for {video_id} via {ssh_host!r}")
+        return None
+    except FileNotFoundError:
+        _log("ssh executable not found; cannot route transcript fetch")
+        return None
+    out = result.stdout or ""
+    if not out.strip().startswith("WEBVTT"):
+        if result.returncode != 0 and result.stderr:
+            first_line = result.stderr.strip().splitlines()[0]
+            _log(
+                f"SSH yt-dlp transcript via {ssh_host!r} failed for "
+                f"{video_id} (rc={result.returncode}): {first_line}"
+            )
+        return None
+    return out
+
+
+def _ytdlp_sub_langs() -> str:
+    """Caption languages to try, from LAST30DAYS_YT_SUB_LANGS (default en,es,pt)."""
+    raw = os.environ.get("LAST30DAYS_YT_SUB_LANGS", "").strip()
+    if not raw:
+        return "en,es,pt"
+    return ",".join(code.strip().lower() for code in raw.split(",") if code.strip()) or "en,es,pt"
+
+
+def _pick_ytdlp_vtt(video_id: str, temp_dir: str, priority: List[str]) -> Optional[Path]:
+    """Return the best on-disk VTT match for video_id, preferring priority order."""
+    matches = list(Path(temp_dir).glob(f"{video_id}*.vtt"))
+    if not matches:
+        return None
+    priority_index = {code: i for i, code in enumerate(priority)}
+
+    def rank(p: Path) -> int:
+        stem = p.stem
+        suffix = stem[len(video_id) + 1:] if stem.startswith(video_id + ".") else ""
+        code = suffix.split("-")[0].split(".")[0]
+        return priority_index.get(code, len(priority_index))
+
+    return sorted(matches, key=rank)[0]
+
+
+def _transcript_backoff(video_id: str, attempt: int) -> float:
+    """Backoff seconds before a transcript retry.
+
+    Staggered per-video (a sub-second offset derived from the id) so parallel
+    workers don't retry in lockstep and re-trip YouTube's limiter.
+    """
+    offset = (sum(ord(c) for c in video_id) % 1000) / 1000.0  # 0.0–1.0s
+    return _TRANSCRIPT_BACKOFF_BASE * (attempt + 1) + offset
+
+
+def _read_vtt(video_id: str, temp_dir: str) -> Optional[str]:
+    """Return the VTT text yt-dlp wrote for ``video_id``, or None if absent."""
+    vtt_path = _pick_ytdlp_vtt(video_id, temp_dir, _ytdlp_sub_langs().split(","))
+    if vtt_path is None:
+        return None
+
+    try:
+        return vtt_path.read_text(encoding="utf-8", errors="replace")
+    except OSError:
+        return None
+
+
+def _fetch_transcript_ytdlp(
+    video_id: str,
+    temp_dir: str,
+    status: Optional[Dict[str, Any]] = None,
+    fast_fail: bool = False,
+) -> Optional[str]:
+    """Fetch transcript using yt-dlp (original implementation).
+
+    Args:
+        video_id: YouTube video ID
+        temp_dir: Temporary directory for subtitle files
+        status: Optional dict mutated to record a yt-dlp failure reason
+            (``status["ytdlp_error"]``) so the caller can tell a real fetch
+            error (rate-limit / bot-check / network / timeout) apart from a
+            video that genuinely has no captions, and skip the misleading
+            "no captions found" log + the YouTube-blocked HTTP fallback.
+        fast_fail: When True, a ScrapeCreators fallback is available, so a
+            transient failure (429 / bot-gate) should fail over fast rather
+            than retry into the same rate limit. Collapses to a single attempt
+            with a shorter per-attempt timeout. The slow thing in a run is
+            yt-dlp retrying, not the fast SC fetch, so this is what keeps
+            "yt-dlp first" from reintroducing the multi-minute hang.
+
+    Returns:
+        Raw VTT text, or None if no captions are available or the fetch failed.
+        On a hard (non-no-caption) failure, sets ``status["ytdlp_error"]``.
+    """
+    cmd = [
+        "yt-dlp",
+        "--ignore-config",
+        "--no-cookies-from-browser",
+        "--write-auto-subs",
+        "--sub-lang", _ytdlp_sub_langs(),
+        "--sub-format", "vtt",
+        "--skip-download",
+        "--no-warnings",
+        "-o", f"{temp_dir}/%(id)s",
+        f"https://www.youtube.com/watch?v={video_id}",
+    ]
+
+    timeout = _TRANSCRIPT_FAST_TIMEOUT if fast_fail else _TRANSCRIPT_TIMEOUT
+    attempts = 1 if fast_fail else _TRANSCRIPT_MAX_RETRIES + 1
+    last_reason: Optional[str] = None
+    for attempt in range(attempts):
+        try:
+            result = subproc.run_with_timeout(cmd, timeout=timeout)
+        except subproc.SubprocTimeout:
+            last_reason = f"timed out after {timeout}s"
+            _log(f"yt-dlp transcript timed out after {timeout}s for {video_id} "
+                 f"(attempt {attempt + 1}/{attempts})")
+            if attempt < attempts - 1:
+                time.sleep(_transcript_backoff(video_id, attempt))
+                continue
+            break
+        except FileNotFoundError:
+            # yt-dlp binary missing — not transient, not retryable.
+            if status is not None:
+                status["ytdlp_error"] = "yt-dlp not found"
+            return None
+
+        if result.returncode == 0:
+            vtt = _read_vtt(video_id, temp_dir)
+            if vtt is not None:
+                return vtt
+            # Exit 0 with no file == the uploader has no matching captions.
+            # Genuine no-captions: return quietly (caller may still try direct).
+            return None
+
+        # Non-zero exit, but yt-dlp may have written a usable VTT before the
+        # failing language errored. With the default `--sub-lang en,es,pt`, an
+        # English video fetches `en` fine, then `es`/`pt` hit a 429 and yt-dlp
+        # exits non-zero — yet the `en` track is already on disk. A partial
+        # success is still a real transcript, so salvage any VTT before
+        # classifying this as an error (and, worse, retrying straight back into
+        # the same rate limit). This is the root cause of the 0/N transcript
+        # runs reported when every video had captions.
+        partial_vtt = _read_vtt(video_id, temp_dir)
+        if partial_vtt is not None:
+            return partial_vtt
+
+        # Non-zero exit == a real error worth classifying & surfacing.
+        stderr = (result.stderr or "").strip()
+        snippet = (stderr.splitlines()[-1][:200] if stderr
+                   else f"exit {result.returncode}")
+        if _NO_CAPTION_RE.search(stderr):
+            # yt-dlp can exit non-zero when the requested language is absent.
+            # Treat as genuine no-captions, not an error worth retrying.
+            return None
+        last_reason = snippet
+        if _TRANSIENT_RE.search(stderr) and attempt < attempts - 1:
+            _log(f"yt-dlp transcript transient failure for {video_id} "
+                 f"(attempt {attempt + 1}/{attempts}): {snippet}")
+            time.sleep(_transcript_backoff(video_id, attempt))
+            continue
+        # Non-transient, or retries exhausted — surface the real reason.
+        _log(f"yt-dlp transcript failed for {video_id} "
+             f"(exit {result.returncode}): {snippet}")
+        break
+
+    if status is not None and last_reason is not None:
+        status["ytdlp_error"] = last_reason
+    return None
+
+
+def _should_try_sc_transcript(status: Optional[Dict[str, Any]]) -> bool:
+    """Whether to spend a ScrapeCreators credit after the keyless cascade failed.
+
+    Skip when the keyless path *proved* the uploader has no caption track
+    (``no_caption_tracks``): SC would also return nothing, so a credit would be
+    wasted. A transient hard failure (``ytdlp_error``: 429 / bot-gate / timeout)
+    is a false negative, so SC is worth trying.
+    """
+    st = status or {}
+    return not st.get("no_caption_tracks")
+
+
+def fetch_transcript(
+    video_id: str,
+    temp_dir: str,
+    status: Optional[Dict[str, Any]] = None,
+    token: Optional[str] = None,
+) -> Optional[str]:
+    """Fetch auto-generated transcript for a YouTube video.
+
+    Uses yt-dlp when available (preferred, more robust). Falls back to
+    direct HTTP transcript fetching when yt-dlp is not installed, and finally
+    to the ScrapeCreators transcript endpoint when a key is present and the
+    keyless cascade comes back empty.
+
+    Args:
+        video_id: YouTube video ID
+        temp_dir: Temporary directory for subtitle files
+        status: Optional dict mutated by the direct-HTTP path to record
+            per-video signals like ``no_caption_tracks``. Used to surface a
+            captions-disabled count so the quality nudge avoids false-positive
+            "stale yt-dlp" flags.
+        token: Optional ScrapeCreators API key. When present, yt-dlp fails over
+            fast (see ``_fetch_transcript_ytdlp`` ``fast_fail``) and a true hard
+            failure falls back to the SC transcript endpoint. A credit is only
+            spent on a genuine yt-dlp failure, never on success and never on a
+            video proven to have no captions. None preserves keyless behavior.
+
+    Returns:
+        Plaintext transcript string, or None if no captions available.
+    """
+    raw_vtt = None
+    ssh_host = _ytdlp_ssh_host()
+    if ssh_host and is_ytdlp_installed():
+        raw_vtt = _fetch_transcript_ytdlp_via_ssh(video_id, ssh_host)
+        if not raw_vtt:
+            _log(f"SSH yt-dlp transcript failed for {video_id}, trying direct HTTP fallback")
+            raw_vtt = _fetch_transcript_direct(video_id, status=status)
+    elif is_ytdlp_installed():
+        raw_vtt = _fetch_transcript_ytdlp(
+            video_id, temp_dir, status=status, fast_fail=bool(token),
+        )
+        if not raw_vtt:
+            ytdlp_error = (status or {}).get("ytdlp_error")
+            if ytdlp_error:
+                # Hard failure (429 / bot-gate / timeout). The direct-HTTP
+                # fallback is also YouTube-blocked, so skip it and let the
+                # ScrapeCreators fallback below handle it when a key is present.
+                _log(f"Transcript fetch failed for {video_id}: {ytdlp_error}")
+            else:
+                _log(f"yt-dlp found no captions for {video_id}, trying direct HTTP fallback")
+                raw_vtt = _fetch_transcript_direct(video_id, status=status)
+    else:
+        _log("yt-dlp not installed, using direct HTTP transcript fetch")
+        raw_vtt = _fetch_transcript_direct(video_id, status=status)
+
+    if raw_vtt:
+        transcript = _clean_vtt(raw_vtt)
+        # Truncate to max words
+        words = transcript.split()
+        if len(words) > TRANSCRIPT_MAX_WORDS:
+            transcript = ' '.join(words[:TRANSCRIPT_MAX_WORDS]) + '...'
+        return transcript if transcript else None
+
+    # Keyless cascade produced nothing. When a ScrapeCreators key is present and
+    # the video was not proven caption-less, fall back to the SC transcript
+    # endpoint (fetched server-side: no 429, cookies, or PO tokens). Returns
+    # already-cleaned, word-capped plaintext.
+    if token and _should_try_sc_transcript(status):
+        sc_transcript = _sc_fetch_transcript(video_id, token)
+        if sc_transcript:
+            return sc_transcript
+
+    _log(f"No transcript available for {video_id}")
+    return None
+
+
+def fetch_transcripts_parallel(
+    video_ids: List[str],
+    max_workers: int = 5,
+    out_captions_disabled: Optional[Set[str]] = None,
+    token: Optional[str] = None,
+) -> Dict[str, Optional[str]]:
+    """Fetch transcripts for multiple videos in parallel.
+
+    Args:
+        video_ids: List of YouTube video IDs
+        max_workers: Max parallel fetches
+        out_captions_disabled: Optional set mutated to record video_ids whose
+            uploader confirmed no caption tracks (vs. transient fetch failures).
+            Backward-compatible: callers that don't care can omit.
+        token: Optional ScrapeCreators API key, threaded to each
+            ``fetch_transcript`` so the per-video SC fallback activates on
+            yt-dlp failure. None preserves keyless behavior.
+
+    Returns:
+        Dict mapping video_id to transcript text (or None).
+    """
+    if not video_ids:
+        return {}
+
+    _log(f"Fetching transcripts for {len(video_ids)} videos")
+
+    results = {}
+    statuses: Dict[str, Dict[str, Any]] = {vid: {} for vid in video_ids}
+    with tempfile.TemporaryDirectory() as temp_dir:
+        with ThreadPoolExecutor(max_workers=max_workers) as executor:
+            futures = {
+                executor.submit(
+                    fetch_transcript, vid, temp_dir, statuses[vid], token,
+                ): vid
+                for vid in video_ids
+            }
+            for future in as_completed(futures):
+                vid = futures[future]
+                try:
+                    results[vid] = future.result()
+                except OSError as exc:
+                    _log(f"Transcript fetch error for {vid}: {exc}")
+                    results[vid] = None
+                except Exception as exc:
+                    _log(f"Unexpected transcript error for {vid}: {type(exc).__name__}: {exc}")
+                    results[vid] = None
+
+    if out_captions_disabled is not None:
+        for vid, st in statuses.items():
+            if st.get("no_caption_tracks"):
+                out_captions_disabled.add(vid)
+
+    got = sum(1 for v in results.values() if v)
+    errors = sum(1 for v in results.values() if v is None)
+    _log(f"Got transcripts for {got}/{len(video_ids)} videos ({errors} failed)")
+    return results
+
+
+def backfill_transcripts(
+    items: List[Any], topic: str = "", depth: str = "default",
+    token: Optional[str] = None,
+) -> None:
+    """Second-pass transcript fetch for finalized items that lack one (#542).
+
+    ``token`` is the optional ScrapeCreators key, threaded to
+    ``fetch_transcripts_parallel`` so the SC fallback covers backfill survivors
+    that yt-dlp can't fetch. None preserves keyless behavior.
+    """
+    limit = TRANSCRIPT_LIMITS.get(depth, TRANSCRIPT_LIMITS["default"])
+    if limit <= 0 or not items or not is_ytdlp_installed():
+        return
+    have = sum(
+        1 for it in items
+        if it.metadata.get("transcript_highlights") or it.metadata.get("transcript_snippet")
+    )
+    need = limit - have
+    if need <= 0:
+        return
+    missing = [
+        it for it in items
+        if it.item_id
+        and not it.metadata.get("transcript_highlights")
+        and not it.metadata.get("transcript_snippet")
+        and not it.metadata.get("captions_disabled")
+    ]
+    attempts = missing[: need * 3]
+    if not attempts:
+        return
+    _log(f"Backfilling transcripts for {len(attempts)} finalized videos (target: {need})")
+    captions_disabled: Set[str] = set()
+    transcripts = fetch_transcripts_parallel(
+        [it.item_id for it in attempts],
+        out_captions_disabled=captions_disabled,
+        token=token,
+    )
+    for it in attempts:
+        if it.item_id in captions_disabled:
+            it.metadata["captions_disabled"] = True
+            continue
+        transcript = transcripts.get(it.item_id)
+        if not transcript:
+            continue
+        it.metadata["transcript_snippet"] = transcript
+        highlights = extract_transcript_highlights(transcript, topic)
+        if highlights:
+            it.metadata["transcript_highlights"] = highlights
+        if not it.snippet:
+            it.snippet = " ".join(transcript.split()[:80])
+
+
+def _transcript_candidate_sort_key(item: dict) -> tuple:
+    """Sort key for transcript candidate selection.
+
+    Combines views with recency so that recent videos (which survive
+    strict_recent freshness pruning) are prioritised over old high-view
+    videos whose transcripts would be discarded downstream.
+    """
+    views = item.get("engagement", {}).get("views", 0) or 0
+    recency = dates.recency_score(item.get("date", ""))
+    return (views, recency)
+
+
+def search_and_transcribe(
+    topic: str,
+    from_date: str,
+    to_date: str,
+    depth: str = "default",
+    token: Optional[str] = None,
+) -> Dict[str, Any]:
+    """Full YouTube search: find videos, then fetch transcripts for top results.
+
+    Uses expand_youtube_queries() to generate multiple search queries,
+    runs yt-dlp for each, and merges/deduplicates results by video ID.
+
+    Args:
+        topic: Search topic
+        from_date: Start date (YYYY-MM-DD)
+        to_date: End date (YYYY-MM-DD)
+        depth: 'quick', 'default', or 'deep'
+        token: Optional ScrapeCreators key for the per-video transcript
+            fallback (threaded to fetch_transcripts_parallel).
+
+    Returns:
+        Dict with 'items' list. Each item has a 'transcript_snippet' field.
+    """
+    # Step 1: Multi-query search — run yt-dlp for each expanded query
+    queries = expand_youtube_queries(topic, depth)
+    seen_ids: Set[str] = set()
+    items: List[Dict[str, Any]] = []
+    for q in queries:
+        search_result = search_youtube(q, from_date, to_date, depth)
+        for item in search_result.get("items", []):
+            vid = item.get("video_id", "")
+            if vid and vid not in seen_ids:
+                seen_ids.add(vid)
+                items.append(item)
+
+    # Sort merged results by views descending
+    items.sort(key=lambda x: x.get("engagement", {}).get("views", 0), reverse=True)
+
+    if not items:
+        return search_result
+
+    # Step 2: Fetch transcripts for top videos.
+    # Sort candidates by a combination of views and recency so that recent
+    # videos (which survive strict_recent pruning) are not starved of
+    # transcript budget by older high-view-count outliers.
+    # Try more candidates than the limit because some videos (music videos,
+    # short clips) lack captions. Attempt up to 3x the limit so we have a
+    # good chance of reaching the target number of successful transcripts.
+    transcript_limit = TRANSCRIPT_LIMITS.get(depth, TRANSCRIPT_LIMITS["default"])
+    transcripts: Dict[str, Optional[str]] = {}
+    captions_disabled_ids: Set[str] = set()
+    if transcript_limit > 0:
+        attempt_count = min(len(items), transcript_limit * 3)
+        transcript_candidates = sorted(
+            items, key=_transcript_candidate_sort_key, reverse=True,
+        )
+        candidate_ids = [item["video_id"] for item in transcript_candidates[:attempt_count]]
+        _log(f"Fetching transcripts for up to {attempt_count} videos (target: {transcript_limit}): {candidate_ids}")
+        transcripts = fetch_transcripts_parallel(
+            candidate_ids, out_captions_disabled=captions_disabled_ids,
+            token=token,
+        )
+        # Record fetch outcomes (captions-disabled videos can never succeed,
+        # so they don't count as failures) for the stale-yt-dlp nudge.
+        _TRANSCRIPT_FETCH_STATS["attempts"] += len(candidate_ids)
+        _TRANSCRIPT_FETCH_STATS["failures"] += sum(
+            1 for vid in candidate_ids
+            if not transcripts.get(vid) and vid not in captions_disabled_ids
+        )
+    else:
+        _log(f"Transcript limit is 0 for depth={depth}, skipping transcript fetch")
+
+    # Step 3: Attach transcripts and extract highlights. Mark captions_disabled
+    # so quality_nudge can subtract those videos from the degraded-ratio
+    # denominator (uploader-disabled captions can never produce a transcript;
+    # counting them was producing false-positive stale-yt-dlp nudges).
+    core_topic = _extract_core_subject(topic)
+    for item in items:
+        vid = item["video_id"]
+        transcript = transcripts.get(vid)
+        item["transcript_snippet"] = transcript or ""
+        item["transcript_highlights"] = extract_transcript_highlights(
+            transcript or "", core_topic,
+        )
+        item["captions_disabled"] = vid in captions_disabled_ids
+
+    return {"items": items}
+
+
+def parse_youtube_response(response: Dict[str, Any]) -> List[Dict[str, Any]]:
+    """Parse YouTube search response to normalized format.
+
+    Returns:
+        List of item dicts ready for normalization.
+    """
+    return response.get("items", [])
+
+
+# ---------------------------------------------------------------------------
+# ScrapeCreators YouTube API support
+# ---------------------------------------------------------------------------
+
+SCRAPECREATORS_YT_BASE = "https://api.scrapecreators.com/v1/youtube"
+
+
+def _total_engagement(item: Dict[str, Any]) -> int:
+    """Combined engagement score for ranking which videos to enrich."""
+    eng = item.get("engagement", {})
+    views = eng.get("views", 0) or 0
+    likes = eng.get("likes", 0) or 0
+    comments = eng.get("comments", 0) or 0
+    return views + likes + comments
+
+
+def enrich_with_comments(
+    items: List[Dict[str, Any]],
+    token: str,
+    max_videos: int = 3,
+    max_comments: int = 5,
+) -> List[Dict[str, Any]]:
+    """Enrich top YouTube videos with comment data from ScrapeCreators.
+
+    For the top N videos by engagement, fetches comments via the SC API
+    and attaches them as a ``top_comments`` field on each item.
+
+    Args:
+        items: YouTube items from search_and_transcribe() or search_youtube_sc()
+        token: ScrapeCreators API key
+        max_videos: How many videos to enrich with comments
+        max_comments: Max comments to keep per video
+
+    Returns:
+        Items list (mutated in place) with top_comments added to enriched items.
+    """
+    if not items or not token or max_videos <= 0:
+        return items
+
+    ranked = sorted(items, key=_total_engagement, reverse=True)
+    top_items = ranked[:max_videos]
+    _log(f"Enriching comments for {len(top_items)} YouTube videos")
+
+    from concurrent.futures import ThreadPoolExecutor, as_completed
+
+    def _enrich_one(item: dict) -> bool:
+        video_id = item.get("video_id", "")
+        if not video_id:
+            return False
+        try:
+            comments = _fetch_video_comments(video_id, token, max_comments)
+            if comments:
+                item["top_comments"] = comments
+                return True
+        except Exception as exc:
+            _log(f"Comment enrichment failed for {video_id}: {exc}")
+        return False
+
+    enriched_count = 0
+    with ThreadPoolExecutor(max_workers=min(4, len(top_items))) as executor:
+        futures = {executor.submit(_enrich_one, item): item for item in top_items}
+        for future in as_completed(futures):
+            if future.result():
+                enriched_count += 1
+
+    _log(f"Enriched {enriched_count}/{len(top_items)} videos with comments")
+    return items
+
+
+def _fetch_video_comments(
+    video_id: str,
+    token: str,
+    max_comments: int = 5,
+) -> List[Dict[str, Any]]:
+    """Fetch comments for a single YouTube video via ScrapeCreators.
+
+    Args:
+        video_id: YouTube video ID
+        token: ScrapeCreators API key
+        max_comments: Maximum comments to return
+
+    Returns:
+        List of comment dicts with author, text, likes, date.
+    """
+    video_url = f"https://www.youtube.com/watch?v={video_id}"
+    try:
+        data = http.get(
+            f"{SCRAPECREATORS_YT_BASE}/video/comments",
+            params={"url": video_url},
+            headers=http.scrapecreators_headers(token),
+            timeout=30,
+            retries=2,
+        )
+    except Exception as exc:
+        _log(f"Comment fetch error for {video_id}: {exc}")
+        return []
+
+    raw_comments = data.get("comments", data.get("data", []))
+    comments = []
+    for c in raw_comments[:max_comments]:
+        text = c.get("text") or c.get("body") or c.get("content", "")
+        if not text:
+            continue
+
+        # SC returns author as {"name": "@handle", ...}; legacy mocks may pass a string.
+        author = c.get("author") or c.get("author_name", "")
+        if isinstance(author, dict):
+            author = author.get("name") or author.get("handle") or ""
+
+        # SC nests likes under engagement.likes; legacy shapes used top-level keys.
+        engagement = c.get("engagement") or {}
+        likes = c.get("likes")
+        if likes is None:
+            likes = engagement.get("likes", 0) if isinstance(engagement, dict) else 0
+        if not likes:
+            likes = c.get("vote_count", 0)
+
+        date = (
+            c.get("date")
+            or c.get("published_at")
+            or c.get("publishedTime")
+            or c.get("publishedTimeText", "")
+        )
+
+        comments.append({
+            "author": author,
+            "text": text[:400],
+            "likes": likes,
+            "date": date,
+        })
+
+    return comments
+
+
+def search_youtube_sc(
+    topic: str,
+    from_date: str,
+    to_date: str,
+    depth: str = "default",
+    token: str = None,
+) -> Dict[str, Any]:
+    """Search YouTube via ScrapeCreators API (fallback when yt-dlp is unavailable).
+
+    Uses SC keyword search to find videos and SC transcript endpoint to
+    fetch transcripts. Called by pipeline.py when yt-dlp fails.
+
+    Args:
+        topic: Search topic
+        from_date: Start date (YYYY-MM-DD)
+        to_date: End date (YYYY-MM-DD)
+        depth: 'quick', 'default', or 'deep'
+        token: ScrapeCreators API key
+
+    Returns:
+        Dict with 'items' list of video metadata dicts.
+    """
+    if not token:
+        return {"items": [], "error": "No SCRAPECREATORS_API_KEY configured"}
+
+    count = DEPTH_CONFIG.get(depth, DEPTH_CONFIG["default"])
+    core_topic = _extract_core_subject(topic)
+    _log(f"Searching YouTube via ScrapeCreators for '{core_topic}' (depth={depth})")
+
+    # Step 1: Search
+    raw_items = _sc_youtube_search(core_topic, token)
+    if not raw_items:
+        _log("SC YouTube search returned 0 results")
+        return {"items": []}
+
+    # Parse into normalized items
+    items = []
+    for i, raw in enumerate(raw_items[:count]):
+        video_id = (
+            raw.get("id") or raw.get("video_id") or raw.get("videoId") or ""
+        )
+        title = raw.get("title", "")
+        channel = raw.get("channel") or raw.get("channel_name") or raw.get("uploader", "")
+        description = str(raw.get("description", ""))[:500]
+        view_count = raw.get("view_count") or raw.get("views", 0)
+        like_count = raw.get("like_count") or raw.get("likes", 0)
+        comment_count = raw.get("comment_count") or raw.get("comments", 0)
+
+        # Date: try multiple field names
+        date_str = raw.get("upload_date") or raw.get("date") or raw.get("published_at", "")
+        if date_str and len(date_str) == 8 and date_str.isdigit():
+            date_str = f"{date_str[:4]}-{date_str[4:6]}-{date_str[6:8]}"
+        elif date_str and "T" in date_str:
+            date_str = date_str[:10]
+
+        url = raw.get("url", "")
+        if not url and video_id:
+            url = f"https://www.youtube.com/watch?v={video_id}"
+
+        items.append({
+            "video_id": video_id,
+            "title": title,
+            "url": url,
+            "channel_name": channel,
+            "date": date_str if date_str else None,
+            "engagement": {
+                "views": view_count or 0,
+                "likes": like_count or 0,
+                "comments": comment_count or 0,
+            },
+            "duration": raw.get("duration"),
+            "relevance": _compute_relevance(core_topic, f"{title} {description}"),
+            "why_relevant": f"YouTube: {title[:60]}" if title else f"YouTube: {core_topic}",
+            "description": description,
+        })
+
+    # Soft date filter
+    recent = [i for i in items if i["date"] and i["date"] >= from_date]
+    if len(recent) >= 3:
+        items = recent
+        _log(f"Found {len(items)} videos within date range")
+    else:
+        _log(f"Found {len(items)} videos ({len(recent)} within date range, keeping all)")
+
+    # Sort by views
+    items.sort(key=lambda x: x["engagement"]["views"], reverse=True)
+
+    # Step 2: Fetch transcripts for top videos
+    transcript_limit = TRANSCRIPT_LIMITS.get(depth, TRANSCRIPT_LIMITS["default"])
+    if transcript_limit > 0 and items:
+        attempt_count = min(len(items), transcript_limit * 3)
+        # Same in-window-first ordering as search_and_transcribe(): don't let
+        # an out-of-window back-catalog (kept by the soft date filter above)
+        # consume the transcript budget of videos the freshness scorer keeps.
+        in_window = [i for i in items if i.get("date") and i["date"] >= from_date]
+        out_of_window = [i for i in items if not (i.get("date") and i["date"] >= from_date)]
+        _log(f"Fetching SC transcripts for up to {attempt_count} videos (target: {transcript_limit})")
+        for item in (in_window + out_of_window)[:attempt_count]:
+            vid = item["video_id"]
+            if not vid:
+                continue
+            transcript = _sc_fetch_transcript(vid, token)
+            item["transcript_snippet"] = transcript or ""
+            item["transcript_highlights"] = extract_transcript_highlights(
+                transcript or "", core_topic,
+            )
+    else:
+        for item in items:
+            item["transcript_snippet"] = ""
+            item["transcript_highlights"] = []
+
+    _log(f"SC YouTube: {len(items)} videos returned")
+    return {"items": items}
+
+
+def _sc_youtube_search(keyword: str, token: str) -> List[Dict[str, Any]]:
+    """Call ScrapeCreators YouTube search endpoint.
+
+    Args:
+        keyword: Search keyword
+        token: ScrapeCreators API key
+
+    Returns:
+        List of raw video dicts from the API.
+    """
+    try:
+        # SC's /v1/youtube/search rejects ?keyword= with HTTP 400; the canonical
+        # parameter for that endpoint is `query`. Other SC endpoints use their
+        # own per-endpoint param names so this was the lone outlier.
+        data = http.get(
+            f"{SCRAPECREATORS_YT_BASE}/search",
+            params={"query": keyword},
+            headers=http.scrapecreators_headers(token),
+            timeout=30,
+            retries=2,
+        )
+        return data.get("videos", data.get("data", data.get("items", [])))
+    except Exception as exc:
+        _log(f"SC YouTube search error: {exc}")
+        return []
+
+
+def _sc_segment_text(seg: Any) -> str:
+    """Extract caption text from a ScrapeCreators transcript segment.
+
+    The transcript endpoint returns a list of segment dicts
+    (``{text, startMs, endMs}``); older/simpler shapes return plain strings.
+    Pull the ``text`` field for dicts so segment metadata is not stringified
+    into the output (``{'text': ...}`` garbage).
+    """
+    if isinstance(seg, dict):
+        # `or ""` (not a get default): a present-but-null `text` returns None,
+        # which would stringify to the literal "None" for silent/music segments.
+        return str(seg.get("text") or "")
+    return str(seg)
+
+
+def _warn_low_sc_credits(data: Dict[str, Any]) -> None:
+    """Surface a low-credit warning from a ScrapeCreators response, if present."""
+    credits = data.get("credits_remaining")
+    if isinstance(credits, (int, float)) and not isinstance(credits, bool):
+        if credits < _SC_LOW_CREDIT_THRESHOLD:
+            _log(f"ScrapeCreators credits low: {int(credits)} remaining "
+                 f"(below {_SC_LOW_CREDIT_THRESHOLD})")
+
+
+def _sc_fetch_transcript(video_id: str, token: str) -> Optional[str]:
+    """Fetch transcript for a YouTube video via ScrapeCreators.
+
+    Args:
+        video_id: YouTube video ID
+        token: ScrapeCreators API key
+
+    Returns:
+        Plaintext transcript string, or None if unavailable.
+    """
+    video_url = f"https://www.youtube.com/watch?v={video_id}"
+    try:
+        data = http.get(
+            f"{SCRAPECREATORS_YT_BASE}/video/transcript",
+            params={"url": video_url},
+            headers=http.scrapecreators_headers(token),
+            timeout=30,
+            retries=1,
+        )
+    except Exception as exc:
+        _log(f"SC transcript error for {video_id}: {exc}")
+        return None
+
+    _warn_low_sc_credits(data)
+
+    transcript = data.get("transcript")
+    if not transcript:
+        return None
+
+    if isinstance(transcript, list):
+        transcript = " ".join(_sc_segment_text(seg) for seg in transcript).strip()
+
+    # Clean VTT formatting if present
+    transcript = _clean_vtt(transcript)
+
+    # Truncate to max words
+    words = transcript.split()
+    if len(words) > TRANSCRIPT_MAX_WORDS:
+        transcript = " ".join(words[:TRANSCRIPT_MAX_WORDS]) + "..."
+
+    return transcript if transcript else None