python ~/.hermes/scripts/portfolio_monthly_review.py

🗂 Directory Structure

~/.hermes/
├── scripts/
│   ├── portfolio_tracker.py
│   ├── portfolio_summary.py
│   └── portfolio_monthly_review.py
└── portfolio/
    ├── portfolio_snapshot.csv     ← latest state
    ├── portfolio_history.csv      ← time-series log
    ├── live_prices.csv            ← your price entries (MANUAL)
    ├── wishlist_conviction.json   ← 1% candidates thesis/exit criteria
    └── reports/
        └── review-2026-04.md      ← monthly markdown reviews

🧑‍💻 Daily / Weekly Workflow

Daily (5 min)

1. Open your broker app / daily report
2. Run: python ~/.hermes/scripts/portfolio_tracker.py --once
   • If prices are stale, run interactive: python …/portfolio_tracker.py
   • Enter last price for each active ticker from your broker 3. Quick peek: python ~/.hermes/scripts/portfolio_summary.py

Monthly (30 min, calendar reminder)

1. Open the latest review: cat ~/.hermes/portfolio/reports/review-*.md
2. Update wishlist_conviction.json:
   • Set "conviction": 1–10 for each planned 1% ticker
   • Refine "thesis", "catalyst", "exit" based on new information 3. Run the review generator: python ~/.hermes/scripts/portfolio_monthly_review.py 4. Commit the report to git / copy to Obsidian 5. Decide capital deployment for the month:
   • Top 1–2 conviction names → new 1% positions
   • Underperforming speculative → trim or exit

⚙️ Automation (cron)

# Add a monthly cron job (first Monday of month at 9 AM)
crontab -e
0 9 1 * * /opt/data/home/.hermes/scripts/portfolio_monthly_review.py

# Pipe to your Hermes report skill (adjust path)
0 9 1 * * /opt/data/home/.hermes/scripts/portfolio_monthly_review.py && python /path/to/kb-report-skill.py --from ~/.hermes/portfolio/reports/review-$(date +\%Y-\%m).md

🎯 Portfolio Design Principles

Current risk profile

• Active holdings: 17 (very diversified)
• Speculative / micro-cap: ~43% of value (ADUR, KRKNF, A6I, ONDS, SIDU, etc.)
• Large-cap core: AMD, GOOGL, AMZN, NFLX, UBER (~45%)
• Mid-cap growth: ZETA, HIMS, OSCR (~30%)

1% position sizing rule

Rebalancing checkpoints

• No speculative > 3% without explicit "core" conviction
• Keep at least 5–10% cash dry powder for dips in ZETA/HIMS/ABCL
• Quarterly: trim anything >2× its target 1% allocation back to 1%

📝 Price Update Tips

Manual entry

ticker,price,updated_at
AMD,329.54,2026-04-29T07:00:00Z
ZETA,18.33,2026-04-29T07:00:00Z

Bulk from broker export

# Convert your broker's CSV to Hermes prices format
import csv
with open("broker_export.csv") as src, open("live_prices.csv","w",newline="") as dst:
    w = csv.writer(dst)
    w.writerow(["ticker","price","updated_at"])
    for row in csv.DictReader(src):
        w.writerow([row["Symbol"], row["Last"], "2026-04-29"])

From your daily Hermes stock news script

# Suppose your daily script outputs TSV: TICKER<tab>PRICE
python ~/.hermes/scripts/daily_stock_prices.py | tee /tmp/todays_prices.tsv
# Then import:
python -c "
import csv, sys
with open('/tmp/todays_prices.tsv') as f, open('~/.hermes/portfolio/live_prices.csv','w',newline='') as out:
    w = csv.writer(out); w.writerow(['ticker','price','updated_at'])
    for line in f:
        t,p = line.strip().split('\\t')
        w.writerow([t, p, 'now'])
"

🔧 Wishlist: Setting Conviction

{
  "RDW": {
    "conviction": 7,
    "max_allocation_pct": 1.0,
    "thesis": "Space infrastructure …",
    "catalyst": "NASA contract …",
    "exit": "Sell if: revenue flat 2 quarters; dilution >10%"
  },
  …
}

🧠 Philosophy

• Data before emotion: Update prices from broker statements, not feelings.
• Thesis-driven: Every position (existing or planned) needs a clear catalyst + exit rule.
• Mechanical sizing: 1% for speculative, 3–5% max for core growth.
• Monthly review cadence: Prevents reactive day-trading; forces strategic thinking.