algo-options

Create an options execution strategy.

Arguments

Parse $ARGUMENTS as: template underlying expiry-date

• $0 = template (short-straddle, iron-condor)
• $1 = underlying (e.g. NIFTY, BANKNIFTY). Default: NIFTY
• $2 = expiry date in DDMMMYY format (e.g. 30DEC25). REQUIRED - no auto-roll. Ask if not given.

If no arguments, ask the user which template.

Instructions

1. Read algo-expert/rules/options-execution.md, unified-strategy-pattern.md, and self-hosted-strategies.md.
2. Confirm with the user that backtest mode is disabled for options strategies (the file will exit with a clear message if --mode backtest is passed). They use OpenAlgo's UI analyzer toggle for sandbox testing instead.
3. Ask the user:
   • Lots: how many lots (default 1)
   • OTM offsets (iron-condor only): OTM_NEAR for short body (default 4), OTM_FAR for long wings (default 8)
   • Entry time (IST, e.g. 09:20)
   • Exit time (IST, e.g. 15:15)
   • SL multiplier (short_straddle, e.g. 1.30 for 30% premium SL)
4. Read the matching template at algo-expert/rules/assets/<template>/strategy.py.
5. Create strategies/<template>_<underlying>/ and copy + customize.
6. Set:
   • UNDERLYING, UNDERLYING_EXCH, EXPIRY_DATE
   • LOTS, LOT_SIZE (NIFTY=65, BANKNIFTY=30, FINNIFTY=60 per Apr 2026 SEBI - see lot-sizes.md)
   • ENTRY_TIME, EXIT_TIME as dtime(H, M)
   • SL_MULTIPLIER for short straddle
   • OTM_NEAR, OTM_FAR for iron condor
7. Tell the user:
   • To upload to /python: requires apscheduler and pytz packages installed (covered by setup)
   • Always re-confirm EXPIRY_DATE before each cycle - the file does NOT auto-roll
   • Set MODE=live in upload form parameters; OpenAlgo UI's analyzer toggle decides sandbox vs real
   • Strategy holds NRML overnight if not flat at EXIT_TIME - cancelallorder + closeposition is called automatically

Templates

Backtest mode is off

The generated file checks for --mode backtest and exits:

Options backtesting is not supported in this skill pack.
Options pricing depends on volatility surfaces, time decay, and OI dynamics
that intraday OHLCV backtests don't capture well.
Use --mode live (with OpenAlgo's UI analyzer toggle for sandbox).

If the user wants to dry-run, they should:

1. Flip OpenAlgo's analyzer toggle ON (in /analyzer UI)
2. Run with --mode live - orders go to the sandbox, not the broker

Avoid

• Do not use icons/emojis
• Do not auto-roll expiry date (EXPIRY_DATE = "30DEC25" is hardcoded; user must update weekly)
• Do not implement options backtesting - explicitly out of scope
• Do not skip the SL placement - per-leg SL is critical for short premium strategies