cd /home/computeruse/work/breaking-monitor && (date -u; ./monitor.py || true; echo '--- tail monitor.log ---'; tail -n 80 monitor.log 2>/dev/null || true) | sed -n '1,200p'

set -euo pipefail
TS=$(date -u +%Y%m%dT%H%M%SZ)
OUT=/home/computeruse/work/day310_final10_${TS}
mkdir -p "$OUT"
UA='breaking-monitor/0.2 (+https://theaidigest.org/village; gpt-5.2@agentvillage.org)'

# Fetch
curl -fsSL -A "$UA" 'https://www.nasdaqtrader.com/rss.aspx?feed=tradehalts' -o "$OUT/nasdaq_tradehalts.xml"
curl -fsSL -A "$UA" 'https://services.swpc.noaa.gov/products/alerts.json' -o "$OUT/swpc_alerts.json"
curl -fsSL -A "$UA" 'https://www.cisa.gov/sites/default/files/feeds/known_exploited_vulnerabilities.json' -o "$OUT/cisa_kev.json"
curl -fsSL -A "$UA" 'https://www.sec.gov/rss/news/press.xml' -o "$OUT/sec_press.xml"
curl -fsSL -A "$UA" 'https://www.nasa.gov/rss/dyn/breaking_news.rss' -o "$OUT/nasa_breaking.xml"

# Hashes
( cd "$OUT" && sha256sum *.xml *.json | sort ) > "$OUT/sha256sums_sorted.txt"

# Parse tops + kev metadata
python - <<'PY'
import json, xml.etree.ElementTree as ET, pathlib, re
out = pathlib.Path("$OUT")

def top_rss(path):
    data = path.read_text(errors='replace')
    # strip leading junk
    m = re.search(r'<rss[^>]*>', data)
    if not m:
        return {"error":"no <rss>"}
    data2 = data[m.start():]
    root = ET.fromstring(data2)
    ch = root.find('channel')
    items = ch.findall('item') if ch is not None else []
    res=[]
    for it in items[:3]:
        def g(tag):
            el=it.find(tag)
            return (el.text or '').strip() if el is not None else ''
        res.append({"title":g('title'),"link":g('link'),"pubDate":g('pubDate'),"guid":g('guid')})
    return {"count":len(items),"top":res}

# SEC
sec = top_rss(out/'sec_press.xml')
# NASA
nasa = top_rss(out/'nasa_breaking.xml')

# SWPC
swpc = json.loads((out/'swpc_alerts.json').read_text())
latest = swpc[0] if swpc else None
swpc_meta = {
  "count": len(swpc),
  "top_product_id": (latest or {}).get('product_id'),
  "top_issue_datetime": (latest or {}).get('issue_datetime'),
  "top_message": (latest or {}).get('message','')[:140]
}

# KEV
kev = json.loads((out/'cisa_kev.json').read_text())
kev_meta = {k: kev.get(k) for k in ['title','dateReleased','count']}
# derive max dateAdded
max_added = None
for v in kev.get('vulnerabilities', []) or []:
    da = v.get('dateAdded')
    if da and (max_added is None or da > max_added):
        max_added = da
kev_meta['max_dateAdded'] = max_added

print('SEC_TOP:', json.dumps(sec, indent=2))
print('NASA_TOP:', json.dumps(nasa, indent=2))
print('SWPC_TOP:', json.dumps(swpc_meta, indent=2))
print('KEV_META:', json.dumps(kev_meta, indent=2))
PY

# NASDAQ structured diff vs watcher state if present
if [ -f /home/computeruse/work/nasdaq_watch_runs/last.xml ]; then
  python /home/computeruse/work/nasdaq_watch.py \
    --date '02/05/2026' \
    --old /home/computeruse/work/nasdaq_watch_runs/last.xml \
    --new "$OUT/nasdaq_tradehalts.xml" \
    > "$OUT/nasdaq_watch_diff.txt" || true
fi

# show diff summary + hashes
echo '--- nasdaq_watch_diff (head) ---'
sed -n '1,120p' "$OUT/nasdaq_watch_diff.txt" 2>/dev/null || true

echo '--- sha256sums_sorted ---'
cat "$OUT/sha256sums_sorted.txt"

echo "OUT=$OUT"

set -euo pipefail
TS=$(date -u +%Y%m%dT%H%M%SZ)
OUT=/home/computeruse/work/day310_final10_${TS}
mkdir -p "$OUT"
UA='breaking-monitor/0.2 (+https://theaidigest.org/village; gpt-5.2@agentvillage.org)'

echo "OUT=$OUT"

fetch() {
  url=$1; out=$2
  echo "FETCH $url"
  curl -fsSL --max-time 25 -A "$UA" "$url" -o "$OUT/$out"
}

fetch 'https://www.nasdaqtrader.com/rss.aspx?feed=tradehalts' nasdaq_tradehalts.xml
fetch 'https://services.swpc.noaa.gov/products/alerts.json' swpc_alerts.json
fetch 'https://www.cisa.gov/sites/default/files/feeds/known_exploited_vulnerabilities.json' cisa_kev.json
fetch 'https://www.sec.gov/rss/news/press.xml' sec_press.xml
fetch 'https://www.nasa.gov/rss/dyn/breaking_news.rss' nasa_breaking.xml

echo OK

set -euo pipefail
OUT=$(ls -1dt /home/computeruse/work/day310_final10_20260205T* | head -n1)
cd "$OUT"
sha256sum *.xml *.json | sort > sha256sums_sorted.txt

python - <<'PY'
import json, re
import xml.etree.ElementTree as ET
from pathlib import Path
out=Path('.')

def top_rss(path: Path):
    txt=path.read_text(errors='replace')
    m=re.search(r'<rss\b[^>]*>', txt)
    if not m:
        return {'error':'no <rss>'}
    txt=txt[m.start():]
    root=ET.fromstring(txt)
    ch=root.find('channel')
    items=ch.findall('item') if ch is not None else []
    top=[]
    for it in items[:5]:
        def g(tag):
            el=it.find(tag)
            return (el.text or '').strip() if el is not None else ''
        top.append({'title':g('title'), 'link':g('link'), 'pubDate':g('pubDate'), 'guid':g('guid')})
    return {'count':len(items), 'top':top}

sec=top_rss(out/'sec_press.xml')
nasa=top_rss(out/'nasa_breaking.xml')

swpc=json.loads((out/'swpc_alerts.json').read_text())
latest=swpc[0] if swpc else {}
swpc_meta={
  'count': len(swpc),
  'top_product_id': latest.get('product_id'),
  'top_issue_datetime': latest.get('issue_datetime'),
  'top_message_prefix': (latest.get('message') or '')[:160],
}

kev=json.loads((out/'cisa_kev.json').read_text())
max_added=None
for v in kev.get('vulnerabilities') or []:
    da=v.get('dateAdded')
    if da and (max_added is None or da>max_added):
        max_added=da
kev_meta={
  'dateReleased': kev.get('dateReleased'),
  'count': kev.get('count'),
  'max_dateAdded': max_added,
}

print('SEC_TOP=', json.dumps(sec, indent=2))
print('NASA_TOP=', json.dumps(nasa, indent=2))
print('SWPC_META=', json.dumps(swpc_meta, indent=2))
print('KEV_META=', json.dumps(kev_meta, indent=2))
PY

cat sha256sums_sorted.txt

set -euo pipefail
NEW_DIR=$(ls -1dt /home/computeruse/work/day310_final10_20260205T* | head -n1)
NEW_XML="$NEW_DIR/nasdaq_tradehalts.xml"
OLD_XML=/home/computeruse/work/day310_latecheck_20260205T205505Z/tradehalts.xml

python /home/computeruse/work/nasdaq_watch.py --date '02/05/2026' --old "$OLD_XML" --new "$NEW_XML" | sed -n '1,200p'

# also diff vs watcher last.xml (may have churn)
if [ -f /home/computeruse/work/nasdaq_watch_runs/last.xml ]; then
  echo '--- vs nasdaq_watch_runs/last.xml ---'
  python /home/computeruse/work/nasdaq_watch.py --date '02/05/2026' --old /home/computeruse/work/nasdaq_watch_runs/last.xml --new "$NEW_XML" | sed -n '1,120p'
fi

ls -la /home/computeruse/work | sed -n '1,200p'
if [ -d /home/computeruse/work/gpt-5-2-news-wire/.git ]; then echo 'repo exists'; else echo 'repo missing'; fi
cd /home/computeruse/work/gpt-5-2-news-wire 2>/dev/null && git status -sb && git log -n 5 --oneline

set -euo pipefail
ls -1 /home/computeruse/work | sed -n '1,200p'
# locate repo
find /home/computeruse/work -maxdepth 2 -type d -name 'gpt-5-2-news-wire' -print

ls -la /home || true
ls -la /home/computeruse || true
ls -la /home/computeruse/work || true
pwd

set +e
ls -la /home
ls -la /home/computeruse
ls -la /home/computeruse/work

set -euo pipefail
cd /home/computeruse/work/gpt-5-2-news-wire

echo 'HEAD:'
git rev-parse HEAD

echo 'Recent commits:'
git log -n 20 --oneline

echo 'Posts:'
ls -1 docs/posts | tail -n 50

echo 'Homepage entries (top 40 lines):'
sed -n '1,120p' docs/index.html | sed -n '1,120p'

set -euo pipefail
cd /home/computeruse/work/gpt-5-2-news-wire

SHAS=(708e4bd9d268b0f1156a21a0407a659f50ae5d2c 82649e9 0f239fb c2ff18f ee057cd 18e9644 692a2d5)
for s in "${SHAS[@]}"; do
  echo "--- $s ---"
  git show -s --format='commit=%H%nsubject=%s%nauthorDate=%aI%ncommitDate=%cI' "$s"
done | sed -n '1,220p'

python - <<'PY'
import urllib.request, json
url='https://api.gdeltproject.org/api/v2/doc/doc?query=NASA%20Selects%20Two%20Earth%20System%20Explorers%20Missions&mode=ArtList&format=json&maxrecords=10&sort=datedesc'
print('fetch', url)
with urllib.request.urlopen(url, timeout=20) as r:
    data=json.load(r)
print('keys', data.keys())
print('count', data.get('totalArticles'))
arts=data.get('articles') or []
for a in arts[:3]:
    print(a.get('seendate'), a.get('domain'), a.get('title')[:80])
PY

URL='https://api.gdeltproject.org/api/v2/doc/doc?query=NASA%20Selects%20Two%20Earth%20System%20Explorers%20Missions&mode=ArtList&format=json&maxrecords=5&sort=datedesc'
echo "$URL"
curl -fsSL --max-time 25 "$URL" | python - <<'PY'
import sys, json
j=json.load(sys.stdin)
print('totalArticles', j.get('totalArticles'))
for a in (j.get('articles') or [])[:5]:
    print(a.get('seendate'), a.get('domain'), a.get('title'))
PY

set -euo pipefail
TMP=/home/computeruse/work/gdelt_tmp.json
URL='https://api.gdeltproject.org/api/v2/doc/doc?query=NASA%20Selects%20Two%20Earth%20System%20Explorers%20Missions&mode=ArtList&format=json&maxrecords=20&sort=datedesc'
curl -fsSL --max-time 25 "$URL" -o "$TMP"
python - <<'PY'
import json
p='/home/computeruse/work/gdelt_tmp.json'
j=json.load(open(p))
print('totalArticles', j.get('totalArticles'))
arts=j.get('articles') or []
for a in arts[:10]:
    print(a.get('seendate'), a.get('domain'), a.get('title'))
PY

set -euo pipefail
URL='https://api.gdeltproject.org/api/v2/doc/doc?query=NASA%20Selects%20Two%20Earth%20System%20Explorers%20Missions&mode=ArtList&format=json&maxrecords=10&sort=datedesc'
OUT=/home/computeruse/work/gdelt_nasa_earth_system_explorers.json

echo "Fetching GDELT..."
# Use aggressive timeouts
curl -fsSL --connect-timeout 5 --max-time 15 "$URL" -o "$OUT" || { echo 'curl failed'; rm -f "$OUT"; exit 0; }

python - <<'PY'
import json
p='/home/computeruse/work/gdelt_nasa_earth_system_explorers.json'
j=json.load(open(p))
print('totalArticles', j.get('totalArticles'))
arts=j.get('articles') or []
for a in arts[:10]:
    print(a.get('seendate'), a.get('domain'), a.get('title'))
PY

set -euo pipefail
python - <<'PY'
import json
p='/home/computeruse/work/gdelt_nasa_earth_system_explorers.json'
j=json.load(open(p))
print('keys:', list(j.keys())[:30])
for k in ['status','message','error','warnings']:
    if k in j:
        print(k, j[k])
# show a short pretty dump (first 1200 chars)
import pprint, textwrap
s=pprint.pformat(j, width=120)
print('dump_prefix:', s[:1200])
PY

# also show raw first 40 lines
sed -n '1,60p' /home/computeruse/work/gdelt_nasa_earth_system_explorers.json

set -euo pipefail
URL='https://api.gdeltproject.org/api/v2/doc/doc?query=NASA%20Selects%20Two%20Earth%20System%20Explorers%20Missions&mode=ArtList&format=json&maxrecords=10&sort=datedesc'
HDR=/home/computeruse/work/gdelt_headers.txt
OUT=/home/computeruse/work/gdelt_body.json
UA='Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120 Safari/537.36'

rm -f "$HDR" "$OUT"
code=$(curl -sS -D "$HDR" -o "$OUT" --connect-timeout 5 --max-time 15 -A "$UA" -H 'Accept: application/json' -w '%{http_code}' "$URL" || echo '000')
echo "http_code=$code"
ls -la "$HDR" "$OUT" || true
sed -n '1,120p' "$HDR" || true
# show first 200 bytes of body
dd if="$OUT" bs=1 count=400 2>/dev/null | sed -e 's/\r/\\r/g' -e 's/\t/\\t/g' -e 's/\n/\\n\n/g'

set -euo pipefail
UA='Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120 Safari/537.36'
URL='https://api.gdeltproject.org/api/v2/doc/doc?query=biden&mode=ArtList&format=json&maxrecords=3&sort=datedesc'
code=$(curl -sS --connect-timeout 5 --max-time 15 -A "$UA" -H 'Accept: application/json' -w '%{http_code}' -o /home/computeruse/work/gdelt_test.json "$URL" || echo '000')
echo "code=$code"
wc -c /home/computeruse/work/gdelt_test.json
cat /home/computeruse/work/gdelt_test.json

set -euo pipefail
UA='Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120 Safari/537.36'
for q in 'Earth System Explorers' 'NASA Selects Two Earth System Explorers' 'Earth System Explorers Missions'; do
  enc=$(python - <<PY
import urllib.parse
print(urllib.parse.quote('''$q'''))
PY
)
  URL="https://api.gdeltproject.org/api/v2/doc/doc?query=${enc}&mode=ArtList&format=json&maxrecords=5&sort=datedesc"
  echo "\n== query: $q =="
  code=$(curl -sS --connect-timeout 5 --max-time 15 -A "$UA" -H 'Accept: application/json' -w '%{http_code}' -o /home/computeruse/work/gdelt_q.json "$URL" || echo '000')
  echo "code=$code size=$(wc -c </home/computeruse/work/gdelt_q.json)"
  python - <<'PY'
import json
j=json.load(open('/home/computeruse/work/gdelt_q.json'))
arts=j.get('articles')
print('has_articles', isinstance(arts,list), 'n', len(arts) if isinstance(arts,list) else None)
if isinstance(arts,list):
  for a in arts[:5]:
    print(a.get('seendate'), a.get('domain'), a.get('title'))
else:
  print('rawkeys', list(j.keys()))
PY

done

cd ~/deepseek-news && pwd && ls -la

find ~ -name "deepseek-news" -type d 2>/dev/null | head -20

cd /home/computeruse/breaking-news/deepseek-news && pwd && ls -la

cd /home/computeruse/breaking-news/deepseek-news && ls -1 _posts | wc -l

cd /home/computeruse/breaking-news/deepseek-news && ls -1 _posts/*2026-02-06* | head -10

cd /home/computeruse/breaking-news/deepseek-news && git log --oneline -20

cd /home/computeruse/breaking-news/deepseek-news && ls -1 _posts/*2026-02-05* | tail -20

cd /home/computeruse/breaking-news/deepseek-news && find _posts -name "*edgar*" -o -name "*sec*" | head -20

cd /home/computeruse/breaking-news/deepseek-news && tail -100 monitor_run.log | grep -i "sec\|edgar\|8-k" | head -20

cd /home/computeruse/breaking-news/deepseek-news && ls -la data/

cd /home/computeruse/breaking-news/deepseek-news && head -30 _posts/2026-02-06-00-10-46-dod_news-844d3f.md

cd /home/computeruse/breaking-news/deepseek-news && grep -n "significance_weights\|source_weights\|weight" monitor_international.py | head -30

cd /home/computeruse/breaking-news/deepseek-news && cat major_news_config.py

cd /home/computeruse/breaking-news/deepseek-news && find _posts -name "*2026-02-05*" -o -name "*2026-02-06*" | wc -l

cd /home/computeruse/breaking-news/deepseek-news && find _posts -name "*sec_edgar*" -o -name "*edgar*" | head -20

cd /home/computeruse/breaking-news/deepseek-news && head -50 _posts/2026-02-05-18-34-12-sec_edgar_batch-151710.md

cd /home/computeruse/breaking-news/deepseek-news && grep -l "8-K" _posts/*2026-02-05* _posts/*2026-02-06* | head -10

cd /home/computeruse/breaking-news/deepseek-news && grep -l "source: SEC EDGAR" _posts/*2026-02-05* _posts/*2026-02-06* | head -10

cd /home/computeruse/breaking-news/deepseek-news && head -100 _posts/2026-02-05-21-10-55-sec_edgar_batch-865c81.md

cd /home/computeruse/breaking-news/deepseek-news && cat > analyze_stories.py << 'EOF'
import os
import re
import glob
import json
from datetime import datetime

def parse_markdown(filepath):
    with open(filepath, 'r', encoding='utf-8') as f:
        content = f.read()
    
    # Parse front matter
    if not content.startswith('---'):
        return None
    parts = content.split('---', 2)
    if len(parts) < 3:
        return None
    front_matter = parts[1]
    
    # Extract fields
    data = {}
    lines = front_matter.strip().split('\n')
    for line in lines:
        if ': ' in line:
            key, value = line.split(': ', 1)
            data[key.strip()] = value.strip()
    
    # Get filename and date from filename pattern
    filename = os.path.basename(filepath)
    # pattern: YYYY-MM-DD-HH-MM-SS-source-hash.md
    match = re.match(r'(\d{4})-(\d{2})-(\d{2})-(\d{2})-(\d{2})-(\d{2})-(.+?)-[a-f0-9]+\.md', filename)
    if match:
        year, month, day, hour, minute, second, source = match.groups()
        data['file_date'] = f"{year}-{month}-{day} {hour}:{minute}:{second}"
        data['source_type'] = source
    else:
        data['file_date'] = data.get('date', '')
        data['source_type'] = 'unknown'
    
    # Extract significance score
    try:
        data['significance'] = float(data.get('significance', 0))
    except:
        data['significance'] = 0
    
    # Extract title
    data['title'] = data.get('title', '')
    
    # Get content preview
    body = parts[2].strip() if len(parts) > 2 else ''
    data['body_preview'] = body[:200] + '...' if len(body) > 200 else body
    
    return data

def main():
    # Get all posts from Feb 5-6 2026
    posts = []
    for filepath in glob.glob('_posts/*2026-02-0[5-6]*.md'):
        data = parse_markdown(filepath)
        if data:
            data['filepath'] = filepath
            posts.append(data)
    
    print(f"Found {len(posts)} posts from Feb 5-6 2026")
    
    # Filter by significance >= 7.0 (major news threshold)
    high_sig = [p for p in posts if p['significance'] >= 7.0]
    print(f"High significance (>=7.0): {len(high_sig)} posts")
    
    # Sort by significance descending
    high_sig.sort(key=lambda x: x['significance'], reverse=True)
    
    # Output top candidates
    print("\n=== TOP CANDIDATES ===")
    for i, p in enumerate(high_sig[:50]):
        print(f"{i+1:3d}. Score: {p['significance']:.2f} | Source: {p['source_type']:20s} | Date: {p['file_date']}")
        print(f"    Title: {p['title'][:80]}")
        print(f"    File: {p['filepath']}")
        print()
    
    # Group by source type
    source_counts = {}
    for p in high_sig:
        src = p['source_type']
        source_counts[src] = source_counts.get(src, 0) + 1
    
    print("\n=== BY SOURCE TYPE ===")
    for src, count in sorted(source_counts.items(), key=lambda x: x[1], reverse=True):
        print(f"{src:25s}: {count:3d}")
    
    # Save to JSON for further review
    with open('candidates_feb5_6.json', 'w') as f:
        json.dump(high_sig, f, indent=2)
    
    print("\nSaved to candidates_feb5_6.json")

if __name__ == '__main__':
    main()
EOF

cd /home/computeruse/breaking-news/deepseek-news && python3 analyze_stories.py 2>&1 | head -100

cd /home/computeruse/breaking-news/deepseek-news && cat _posts/2026-02-05-21-10-55-sec_edgar_batch-865c81.md

cd /home/computeruse/breaking-news/deepseek-news && python3 -c "
import glob, os, re
sources = {}
for f in glob.glob('_posts/*2026-02-05*.md'):
    m = re.match(r'.*?-(\w+)-[a-f0-9]+\.md', os.path.basename(f))
    if m:
        src = m.group(1)
        sources[src] = sources.get(src, 0) + 1
for src, cnt in sorted(sources.items(), key=lambda x: x[1], reverse=True):
    print(f'{src}: {cnt}')
"

cd /home/computeruse/breaking-news/deepseek-news && cat _posts/2026-02-05-21-12-49-prnewswire-a0b743.md | head -30

cd /home/computeruse/breaking-news/deepseek-news && cat _posts/2026-02-05-21-16-48-us_navy-de9873.md | head -30

cd /home/computeruse/breaking-news/deepseek-news && python3 -c "
import glob, os, re, datetime

def parse_md(fpath):
    with open(fpath, 'r', encoding='utf-8') as f:
        text = f.read()
    if not text.startswith('---'):
        return None
    parts = text.split('---', 2)
    if len(parts) < 3:
        return None
    front = parts[1]
    data = {}
    for line in front.strip().split('\\n'):
        if ': ' in line:
            k,v = line.split(': ', 1)
            data[k.strip()] = v.strip()
    try:
        data['significance'] = float(data.get('significance', 0))
    except:
        data['significance'] = 0
    # get source from filename
    basename = os.path.basename(fpath)
    m = re.match(r'.*?-(\w+)-[a-f0-9]+\.md', basename)
    data['source_type'] = m.group(1) if m else 'unknown'
    # get file timestamp
    m2 = re.match(r'(\\d{4})-(\\d{2})-(\\d{2})-(\\d{2})-(\\d{2})-(\\d{2})', basename)
    if m2:
        y,mo,d,h,mi,s = map(int, m2.groups())
        data['file_dt'] = datetime.datetime(y,mo,d,h,mi,s)
    else:
        data['file_dt'] = None
    data['title'] = data.get('title', '')
    data['url'] = data.get('source_url', '')
    return data

candidates = []
for f in glob.glob('_posts/*2026-02-05*.md'):
    d = parse_md(f)
    if d and d['significance'] >= 8.0 and d['source_type'] != 'sec_edgar_batch':
        candidates.append(d)

print(f'Found {len(candidates)} candidates with significance >=8.0 (excluding sec_edgar_batch)')
# sort by significance desc
candidates.sort(key=lambda x: x['significance'], reverse=True)
for i, c in enumerate(candidates[:30]):
    print(f\"{i+1:2d}. {c['significance']:.2f} {c['source_type']:15s} {c['file_dt']} {c['title'][:80]}\")
" 2>&1

cd /home/computeruse/breaking-news/deepseek-news && grep -l "cisa_kev" _posts/*2026-02-05*.md

cd /home/computeruse/breaking-news/deepseek-news && python3 -c "
import glob, os, re
for f in glob.glob('_posts/*2026-02-05*.md'):
    m = re.match(r'.*?-(\w+)-[a-f0-9]+\.md', os.path.basename(f))
    if m and m.group(1) == 'cisa_kev':
        print(f)
        with open(f, 'r') as fp:
            lines = fp.readlines()
            for line in lines[:10]:
                if 'title' in line or 'significance' in line:
                    print(line.strip())
        print()
"

cd /home/computeruse/breaking-news/deepseek-news && python3 -c "
import glob, os, re
for f in glob.glob('_posts/*2026-02-05*.md'):
    m = re.match(r'.*?-(\w+)-[a-f0-9]+\.md', os.path.basename(f))
    if m and m.group(1) == 'nasdaq_halt':
        with open(f, 'r') as fp:
            text = fp.read()
        if 'significance: 7.00' in text:  # they're all 7.0
            parts = text.split('---', 2)
            if len(parts) >= 2:
                front = parts[1]
                for line in front.strip().split('\\n'):
                    if 'title' in line:
                        print(line.strip())
                        print(f)
                        break
" | head -5

cd /home/computeruse/breaking-news/deepseek-news && cat _posts/2026-02-05-21-12-54-prnewswire-90aeec.md | head -40

cd /home/computeruse/breaking-news/deepseek-news && grep -l "nasa_breaking" _posts/*2026-02-05*.md

cd /home/computeruse/breaking-news/deepseek-news && git log --oneline --grep="Illumina" --all | head -5

cd /home/computeruse/breaking-news/deepseek-news && git log --oneline --since="2026-02-05" -- _posts/2026-02-05-21-12-49-prnewswire-a0b743.md | head -5

cd /home/computeruse/breaking-news/deepseek-news && cat > select_top5.py << 'EOF'
import os
import re
import glob
import json
from datetime import datetime, timezone

def parse_md(filepath):
    with open(filepath, 'r', encoding='utf-8') as f:
        content = f.read()
    if not content.startswith('---'):
        return None
    parts = content.split('---', 2)
    if len(parts) < 3:
        return None
    front = parts[1]
    data = {}
    for line in front.strip().split('\n'):
        if ': ' in line:
            k, v = line.split(': ', 1)
            data[k.strip()] = v.strip()
    try:
        data['significance'] = float(data.get('significance', 0))
    except:
        data['significance'] = 0
    # get source type from filename
    basename = os.path.basename(filepath)
    m = re.match(r'.*?-(\w+)-[a-f0-9]+\.md', basename)
    data['source_type'] = m.group(1) if m else 'unknown'
    # get file timestamp
    m2 = re.match(r'(\d{4})-(\d{2})-(\d{2})-(\d{2})-(\d{2})-(\d{2})', basename)
    if m2:
        y, mo, d, h, mi, s = map(int, m2.groups())
        data['file_dt'] = datetime(y, mo, d, h, mi, s, tzinfo=timezone.utc)
    else:
        data['file_dt'] = None
    data['title'] = data.get('title', '')
    data['url'] = data.get('source_url', '')
    data['filename'] = basename
    data['filepath'] = filepath
    return data

def assign_category(data):
    title = data['title'].lower()
    source = data['source_type']
    # categorize
    if source == 'sec_edgar_batch':
        if '8-k' in title.lower():
            return 'sec_8k_material_event'
        elif '10-k' in title.lower() or '10-q' in title.lower():
            return 'sec_earnings_filing'
        else:
            return 'sec_other_filing'
    elif source == 'prnewswire':
        if any(term in title for term in ['earnings', 'results', 'revenue', 'profit', 'quarter']):
            return 'corporate_earnings'
        elif any(term in title for term in ['raises', 'funding', 'investment', '$', 'million', 'billion']):
            return 'funding_round'
        elif any(term in title for term in ['merger', 'acquisition', 'acquires', 'buys']):
            return 'merger_acquisition'
        elif any(term in title for term in ['lawsuit', 'settlement', 'verdict', 'suit']):
            return 'legal_action'
        else:
            return 'corporate_announcement'
    elif source in ('us_navy', 'us_army', 'dod_news'):
        return 'defense_military'
    elif source == 'cisa_kev':
        return 'cybersecurity_vulnerability'
    elif source == 'federal_register':
        return 'government_regulation'
    elif source == 'nasa_breaking':
        return 'space_science'
    elif source == 'who_news':
        return 'health_who'
    else:
        return 'other'

def score_potential_spread(data, category):
    # heuristic for likely media coverage
    base = data['significance']
    # adjust based on category
    if category in ('sec_8k_material_event', 'corporate_earnings', 'merger_acquisition'):
        base *= 1.5  # high financial media interest
    elif category in ('defense_military', 'cybersecurity_vulnerability'):
        base *= 1.3  # specialized but important
    elif category in ('government_regulation', 'space_science'):
        base *= 1.2  # niche but can be big
    # check for major company names
    title = data['title'].lower()
    major_companies = ['mastercard', 'amazon', 'jpmorgan', 'oracle', 'illumina', 
                       'cleanspark', 'philip morris', 'alphabet', 'google']
    for comp in major_companies:
        if comp in title:
            base *= 1.4
            break
    return base

def main():
    posts = []
    for filepath in glob.glob('_posts/*2026-02-0[5-6]*.md'):
        data = parse_md(filepath)
        if data and data['significance'] >= 7.0:
            data['category'] = assign_category(data)
            data['spread_score'] = score_potential_spread(data, data['category'])
            posts.append(data)
    
    print(f"Analyzing {len(posts)} high-significance posts from Feb 5-6")
    
    # Sort by spread score
    posts.sort(key=lambda x: x['spread_score'], reverse=True)
    
    # Output top 20
    print("\n=== TOP 20 CANDIDATES BY SPREAD POTENTIAL ===")
    for i, p in enumerate(posts[:20]):
        print(f"{i+1:2d}. Spread: {p['spread_score']:.2f} | Sig: {p['significance']:.2f} | "
              f"Source: {p['source_type']:15s} | Cat: {p['category']:25s}")
        print(f"    Title: {p['title'][:80]}")
        print(f"    Date: {p['file_dt']}")
        print()
    
    # Group by category
    cats = {}
    for p in posts[:50]:
        cat = p['category']
        cats[cat] = cats.get(cat, 0) + 1
    
    print("\n=== CATEGORY DISTRIBUTION (top 50) ===")
    for cat, cnt in sorted(cats.items(), key=lambda x: x[1], reverse=True):
        print(f"{cat:30s}: {cnt:3d}")
    
    # Select top 5, trying to diversify categories
    selected = []
    seen_cats = set()
    for p in posts:
        if p['category'] not in seen_cats:
            selected.append(p)
            seen_cats.add(p['category'])
        if len(selected) >= 5:
            break
    
    # If not enough categories, fill with highest spread
    if len(selected) < 5:
        for p in posts:
            if p not in selected:
                selected.append(p)
            if len(selected) >= 5:
                break
    
    print("\n=== PROPOSED TOP 5 (DIVERSE CATEGORIES) ===")
    for i, p in enumerate(selected):
        print(f"\n{i+1}. {p['title']}")
        print(f"   Category: {p['category']}")
        print(f"   Source: {p['source_type']}")
        print(f"   Significance: {p['significance']:.2f}")
        print(f"   Spread Score: {p['spread_score']:.2f}")
        print(f"   Published: {p['file_dt']}")
        print(f"   URL: {p['url']}")
        print(f"   File: {p['filename']}")
    
    # Save selection for webpage generation
    with open('top5_selection.json', 'w') as f:
        json.dump([{k: v for k, v in p.items() if k != 'filepath'} for p in selected], f, indent=2, default=str)
    
    print("\nSaved to top5_selection.json")

if __name__ == '__main__':
    main()
EOF

cd /home/computeruse/breaking-news/deepseek-news && python3 select_top5.py 2>&1 | head -80

cd /home/computeruse/breaking-news/deepseek-news && python3 select_top5.py 2>&1 | tail -60

cd /home/computeruse/breaking-news/deepseek-news && cat index.md

cd /home/computeruse/breaking-news/deepseek-news && cat > get_commits.py << 'EOF'
import json
import subprocess
import os

def get_git_commit_for_file(filepath):
    try:
        cmd = ['git', 'log', '--oneline', '--', filepath]
        result = subprocess.run(cmd, capture_output=True, text=True, check=True)
        lines = result.stdout.strip().split('\n')
        if lines:
            first_line = lines[0]
            # format: "hash message"
            parts = first_line.split(' ', 1)
            if len(parts) >= 1:
                return {
                    'hash': parts[0],
                    'message': parts[1] if len(parts) > 1 else ''
                }
    except Exception as e:
        print(f"Error getting commit for {filepath}: {e}")
    return None

def main():
    with open('top5_selection.json', 'r') as f:
        stories = json.load(f)
    
    for story in stories:
        filename = story['filename']
        filepath = os.path.join('_posts', filename)
        commit = get_git_commit_for_file(filepath)
        if commit:
            story['commit_hash'] = commit['hash']
            story['commit_message'] = commit['message']
            # also get timestamp
            cmd = ['git', 'show', '--no-patch', '--format=%cI', commit['hash']]
            result = subprocess.run(cmd, capture_output=True, text=True, check=True)
            story['commit_timestamp'] = result.stdout.strip()
        else:
            story['commit_hash'] = 'unknown'
            story['commit_message'] = 'unknown'
            story['commit_timestamp'] = 'unknown'
    
    with open('top5_with_commits.json', 'w') as f:
        json.dump(stories, f, indent=2)
    
    print("Added commit info to stories")
    for story in stories:
        print(f"{story['title'][:60]}...")
        print(f"  Commit: {story.get('commit_hash', 'N/A')} at {story.get('commit_timestamp', 'N/A')}")
        print()

if __name__ == '__main__':
    main()
EOF

cd /home/computeruse/breaking-news/deepseek-news && python3 get_commits.py 2>&1

cd /home/computeruse/breaking-news/deepseek-news && git status 2>&1 | head -20

cd /home/computeruse/breaking-news/deepseek-news && pwd