Model

Compliance officer: 20 ways to use Ora

Surveillance context: unusual moves around events, restricted lists and documented checks. Each example is a complete, runnable request with the real answers it returned (44 requests, captured 2026-10-31), what they mean and, where it fits, the action it would send.

Run withAnswers were captured on ora-fin-0.1.1. Another model changes the request JSON and the Playground link; run it there to see its answers.

Rule functions and facts used by this player (51)

offset 10 avg 8 ref 7 abs 5 sma 5 atr 4 days_to_earnings 3 highest 3 max 3 zscore 2 days_since_earnings 2 days_to_ex_div 2 pays_dividend 2 atr_pct 2 log 2 sqrt 2 std 2 percentrank 1 gap_down 1 gap_up 1 beat_eps_last 1 beat_revenue_last 1 eps_surprise_pct 1 last_earnings_gap 1 last_earnings_move 1 days_since_ex_div 1 div_yield_ttm 1 days_since_reverse_split 1 days_since_split 1 days_to_split 1 last_split_ratio 1 marubozu 1 outside_bar 1 macd 1 macdsignal 1 rsi 1 next_report_amc 1 next_report_bmo 1 median 1 roc 1 three_white_soldiers 1 doji 1 div_ttm 1 last_div_amount 1 next_div_amount 1 barssince 1 crossdown 1 crossup 1 lowest 1 engulfing_bull 1 hammer 1

Examples 1–10 of 20 · page 1 of 2

Compliance officer · example 1

NVDA: document unusual volume ahead of an earnings report

When volume spikes in the days before a scheduled report, record the market context in the surveillance case system for review.

ora-fin-0.1.1NVDA5 questions2 actions

StateNVDA
Questions5 asked
Answersreal, captured
Actions2 rules · dry run

State

{"symbol":"NVDA"} The server fills in the latest completed session.

Questions

`earn`	rule	`days_to_earnings <= 10` Reports earnings within 10 days
`spike`	rule	`V > 2 * ref(avg(V, 20), -1)` Volume more than twice the previous 20-day average
`rank`	rule	`percentrank(V, 50) > 90` Volume in the top tenth of 50 sessions
`big`	rule	`abs(C - C[-1]) > 1.5 * atr(14)` Today's move was bigger than 1.5 ATR
`spike_tom`	forecast	Will volume spike tomorrow? A volume spike tomorrow

Answers

Question	Answer	Reading
`earn` Reports earnings within 10 days	0.0563	false
`spike` Volume more than twice the previous 20-day average	0.1794	false
`rank` Volume in the top tenth of 50 sessions	0.5398	on the line (true)
`big` Today's move was bigger than 1.5 ATR	0.4404	on the line (false)
`spike_tom` A volume spike tomorrow	0.044	4% · unlikely

Actions

Rules on the answers decide what to send.

earn >= 0.6 and (spike >= 0.6 or rank >= 0.6)→Webhooksurveillance-cases

else→Slackcompliance-log

What each action would send 1 of 2 fired · dry run, nothing sent

caserule false: skipped

logwould send (dry run)Slack → compliance-log

{
  "text": "NVDA (2026-10-30): pre-earnings volume check run; no elevated volume (spike 18%, rank 54%)."
}

What it means

For NVDA on the session of 2026-10-30: Not true: Reports earnings within 10 days; Volume more than twice the previous 20-day average. On the line: Volume in the top tenth of 50 sessions; Today's move was bigger than 1.5 ATR. Forecasts: A volume spike tomorrow 4%. Fired: log (Slack → compliance-log).

The case record holds market context for a reviewer; it is not a finding about anyone, and the else branch documents that the check ran.

Request JSON

{
  "model": "ora-fin-0.1.1",
  "state": {
    "symbol": "NVDA"
  },
  "questions": {
    "earn": {
      "type": "yes_no",
      "ask": "days_to_earnings <= 10"
    },
    "spike": {
      "type": "yes_no",
      "ask": "V > 2 * ref(avg(V, 20), -1)"
    },
    "rank": {
      "type": "yes_no",
      "ask": "percentrank(V, 50) > 90"
    },
    "big": {
      "type": "yes_no",
      "ask": "abs(C - C[-1]) > 1.5 * atr(14)"
    },
    "spike_tom": {
      "type": "yes_no",
      "instructions": "Will volume spike tomorrow?"
    }
  },
  "actions": [
    {
      "id": "case",
      "when": "earn >= 0.6 and (spike >= 0.6 or rank >= 0.6)",
      "do": "webhook",
      "connection": "surveillance-cases",
      "body": {
        "symbol": "{{symbol}}",
        "as_of": "{{as_of}}",
        "type": "market_context",
        "reason": "elevated_volume_before_scheduled_earnings",
        "facts": {
          "earnings_within_10d": "{{earn}}",
          "volume_2x_avg": "{{spike}}",
          "volume_rank_90": "{{rank}}",
          "move_over_1_5_atr": "{{big}}"
        },
        "note": "Context only; no finding."
      }
    },
    {
      "id": "log",
      "when": "else",
      "do": "slack",
      "connection": "compliance-log",
      "message": "{{symbol}} ({{as_of}}): pre-earnings volume check run; no elevated volume (spike {{spike|pct}}, rank {{rank|pct}})."
    }
  ],
  "dry_run": true
}

Response JSON (NVDA, real)

{
  "model": "ora-fin-0.1.1",
  "answers": {
    "earn": {
      "yes_no": 0.0563,
      "rule": "days_to_earnings <= 10",
      "basis": "rule"
    },
    "spike": {
      "yes_no": 0.1794,
      "rule": "V > 2 * ref(avg(V, 20), -1)",
      "basis": "rule"
    },
    "rank": {
      "yes_no": 0.5398,
      "rule": "percentrank(V, 50) > 90",
      "basis": "rule"
    },
    "big": {
      "yes_no": 0.4404,
      "rule": "abs(C - C[-1]) > 1.5 * atr(14)",
      "basis": "rule"
    },
    "spike_tom": {
      "yes_no": 0.044
    }
  },
  "context": {
    "symbol": "NVDA",
    "as_of": "2026-10-30"
  },
  "actions": [
    {
      "id": "case",
      "fired": false,
      "status": "skipped"
    },
    {
      "id": "log",
      "fired": true,
      "status": "dry_run",
      "do": "slack",
      "connection": "compliance-log",
      "preview": {
        "text": "NVDA (2026-10-30): pre-earnings volume check run; no elevated volume (spike 18%, rank 54%)."
      }
    }
  ]
}

abs atr avg days_to_earnings offset percentrank ref

Try it in the Playground

Compliance officer · example 2

Restricted-list names: large moves to note in the daily log

Each evening, check names on the restricted list for large moves or shock days and email the log to compliance.

ora-fin-0.1.14 symbols, one request each4 questions1 actionworkflow · Mon–Fri at 17:15 ET

ScheduleMon–Fri at 17:15 ET
State4 symbols
Questions4 asked
Answersreal, captured
Actions1 rule · dry run

State

{"symbol":"JPM"} then GS, BAC, C, one request each. The server fills in the latest completed session.

Questions

`big`	rule	`abs(C - C[-1]) > 1.5 * atr(14)` Today's move was bigger than 1.5 ATR
`shock`	rule	`zscore(C / C[-1] - 1, 20) < -2` A two-sigma down day
`vol`	rule	`V > 2 * ref(avg(V, 20), -1)` Volume more than twice the previous 20-day average
`gap`	rule	`abs(O - C[-1]) > 0.5 * ref(atr(14), -1)` Gap bigger than half an ATR

Answers

Symbol	`big`rule	`shock`rule	`vol`rule	`gap`rule	Fired
`JPM`	0.101	0.5937	0.0832	0.153	none
`GS`	0.2824	0.6964	0.0537	0.9104	`log`
`BAC`	0.1309	0.2192	0.061	0.1685	none
`C`	0.1024	0.1851	0.1207	0.6901	none

true false not known for rules; forecasts are odds. Hover a column for its question.

Actions

Rules on the answers decide what to send. Each symbol runs its own request, so each one can fire.

big >= 0.6 or shock >= 0.6 or vol >= 0.6→Emailcompliance-team

JPM 0 of 1 fired · dry run, nothing sent

GS 1 of 1 fired · dry run, nothing sent

logwould send (dry run)Email → compliance-team

{
  "subject": "Restricted list: GS large move 2026-10-30",
  "text": "GS: move over 1.5 ATR 28%, return shock 70%, volume spike 5%, gap vs ATR 91%. Review employee and firm activity per procedure."
}

BAC 0 of 1 fired · dry run, nothing sent

C 0 of 1 fired · dry run, nothing sent

What it means

Session of 2026-10-30, 4 symbols: Today's move was bigger than 1.5 ATR: true for none of 4. A two-sigma down day: true for GS of 4. Volume more than twice the previous 20-day average: true for none of 4. Gap bigger than half an ATR: true for GS, C of 4. log fired for GS.

Large moves in restricted names trigger the standard review procedure; quiet days send nothing.

As a scheduled workflow (Mon–Fri at 17:15 ET)

Create it under Workflows in the console: it runs these questions for every symbol on the schedule and fires the actions per symbol. The answers above are what one run returned.

{
  "name": "Restricted-list names: large moves to note in the daily log",
  "symbols": "JPM GS BAC C",
  "days": [
    1,
    2,
    3,
    4,
    5
  ],
  "at_time": "17:15",
  "model": "ora-fin-0.1.1",
  "questions": {
    "big": {
      "type": "yes_no",
      "ask": "abs(C - C[-1]) > 1.5 * atr(14)"
    },
    "shock": {
      "type": "yes_no",
      "ask": "zscore(C / C[-1] - 1, 20) < -2"
    },
    "vol": {
      "type": "yes_no",
      "ask": "V > 2 * ref(avg(V, 20), -1)"
    },
    "gap": {
      "type": "yes_no",
      "ask": "abs(O - C[-1]) > 0.5 * ref(atr(14), -1)"
    }
  },
  "actions": [
    {
      "id": "log",
      "when": "big >= 0.6 or shock >= 0.6 or vol >= 0.6",
      "do": "email",
      "connection": "compliance-team",
      "subject": "Restricted list: {{symbol}} large move {{as_of}}",
      "message": "{{symbol}}: move over 1.5 ATR {{big|pct}}, return shock {{shock|pct}}, volume spike {{vol|pct}}, gap vs ATR {{gap|pct}}. Review employee and firm activity per procedure."
    }
  ]
}

Request JSON (JPM; the others differ only in the symbol)

{
  "model": "ora-fin-0.1.1",
  "state": {
    "symbol": "JPM"
  },
  "questions": {
    "big": {
      "type": "yes_no",
      "ask": "abs(C - C[-1]) > 1.5 * atr(14)"
    },
    "shock": {
      "type": "yes_no",
      "ask": "zscore(C / C[-1] - 1, 20) < -2"
    },
    "vol": {
      "type": "yes_no",
      "ask": "V > 2 * ref(avg(V, 20), -1)"
    },
    "gap": {
      "type": "yes_no",
      "ask": "abs(O - C[-1]) > 0.5 * ref(atr(14), -1)"
    }
  },
  "actions": [
    {
      "id": "log",
      "when": "big >= 0.6 or shock >= 0.6 or vol >= 0.6",
      "do": "email",
      "connection": "compliance-team",
      "subject": "Restricted list: {{symbol}} large move {{as_of}}",
      "message": "{{symbol}}: move over 1.5 ATR {{big|pct}}, return shock {{shock|pct}}, volume spike {{vol|pct}}, gap vs ATR {{gap|pct}}. Review employee and firm activity per procedure."
    }
  ],
  "dry_run": true
}

Response JSON (JPM, real)

{
  "model": "ora-fin-0.1.1",
  "answers": {
    "big": {
      "yes_no": 0.101,
      "rule": "abs(C - C[-1]) > 1.5 * atr(14)",
      "basis": "rule"
    },
    "shock": {
      "yes_no": 0.5937,
      "rule": "zscore(C / C[-1] - 1, 20) < -2",
      "basis": "rule"
    },
    "vol": {
      "yes_no": 0.0832,
      "rule": "V > 2 * ref(avg(V, 20), -1)",
      "basis": "rule"
    },
    "gap": {
      "yes_no": 0.153,
      "rule": "abs(O - C[-1]) > 0.5 * ref(atr(14), -1)",
      "basis": "rule"
    }
  },
  "context": {
    "symbol": "JPM",
    "as_of": "2026-10-30"
  },
  "actions": [
    {
      "id": "log",
      "fired": false,
      "status": "skipped"
    }
  ]
}

abs atr avg offset ref zscore

Try it in the Playground

Compliance officer · example 3

TSLA gap documentation for the surveillance file

Record the size and character of an opening gap so later reviews of trading around it have the facts.

ora-fin-0.1.1TSLA7 questions

StateTSLA
Questions7 asked
Answersreal, captured
Actionsnone

State

{"symbol":"TSLA"} The server fills in the latest completed session.

Questions

`gap_up`	rule	`O > C[-1] * 1.01` Opened more than 1% above yesterday's close
`gap_dn`	rule	`O < C[-1] * 0.99` Opened more than 1% below yesterday's close
`gap_atr`	rule	`abs(O - C[-1]) > 0.5 * ref(atr(14), -1)` Gap bigger than half an ATR
`unfilled`	English → rule	Did the stock gap up with today's low above yesterday's high?
`faded`	rule	`O > C[-1] and C < C[-1]` Gapped up but closed under yesterday's close
`gap_up_bar`	rule	`gap_up()` gap up candle: Today's whole range is above yesterday's high: an unfilled gap up.
`gap_dn_bar`	rule	`gap_down()` gap down candle: Today's whole range is below yesterday's low.

Answers

Question	Answer	Reading
`gap_up` Opened more than 1% above yesterday's close	0.0562	false
`gap_dn` Opened more than 1% below yesterday's close	0.8793	true
`gap_atr` Gap bigger than half an ATR	0.8646	true
`unfilled` Did the stock gap up with today's low above yesterday's high compiled to `gap_up()`	0.0577	false
`faded` Gapped up but closed under yesterday's close	0.0708	false
`gap_up_bar` gap up candle: Today's whole range is above yesterday's high: an unfilled gap up.	0.0577	false
`gap_dn_bar` gap down candle: Today's whole range is below yesterday's low.	0.2768	false

What it means

For TSLA on the session of 2026-10-30: True now: Opened more than 1% below yesterday's close; Gap bigger than half an ATR. Not true: Opened more than 1% above yesterday's close; Did the stock gap up with today's low above yesterday's high; Gapped up but closed under yesterday's close; gap up candle: Today's whole range is above yesterday's high: an unfilled gap up.; gap down candle: Today's whole range is below yesterday's low..

Gap facts are written to the file as they are; reviewers use them to judge whether activity before the gap needs attention.

Request JSON

{
  "model": "ora-fin-0.1.1",
  "state": {
    "symbol": "TSLA"
  },
  "questions": {
    "gap_up": {
      "type": "yes_no",
      "ask": "O > C[-1] * 1.01"
    },
    "gap_dn": {
      "type": "yes_no",
      "ask": "O < C[-1] * 0.99"
    },
    "gap_atr": {
      "type": "yes_no",
      "ask": "abs(O - C[-1]) > 0.5 * ref(atr(14), -1)"
    },
    "unfilled": {
      "type": "yes_no",
      "instructions": "Did the stock gap up with today's low above yesterday's high?"
    },
    "faded": {
      "type": "yes_no",
      "ask": "O > C[-1] and C < C[-1]"
    },
    "gap_up_bar": {
      "type": "yes_no",
      "ask": "gap_up()"
    },
    "gap_dn_bar": {
      "type": "yes_no",
      "ask": "gap_down()"
    }
  }
}

Response JSON (TSLA, real)

{
  "model": "ora-fin-0.1.1",
  "answers": {
    "gap_up": {
      "yes_no": 0.0562,
      "rule": "O > C[-1] * 1.01",
      "basis": "rule"
    },
    "gap_dn": {
      "yes_no": 0.8793,
      "rule": "O < C[-1] * 0.99",
      "basis": "rule"
    },
    "gap_atr": {
      "yes_no": 0.8646,
      "rule": "abs(O - C[-1]) > 0.5 * ref(atr(14), -1)",
      "basis": "rule"
    },
    "unfilled": {
      "yes_no": 0.0577,
      "rule": "gap_up()",
      "basis": "pattern"
    },
    "faded": {
      "yes_no": 0.0708,
      "rule": "O > C[-1] and C < C[-1]",
      "basis": "rule"
    },
    "gap_up_bar": {
      "yes_no": 0.0577,
      "rule": "gap_up()",
      "basis": "rule"
    },
    "gap_dn_bar": {
      "yes_no": 0.2768,
      "rule": "gap_down()",
      "basis": "rule"
    }
  },
  "context": {
    "symbol": "TSLA",
    "as_of": "2026-10-30"
  }
}

abs atr gap_down gap_up offset ref

Try it in the Playground

Compliance officer · example 4

JPM earnings reaction: context for the trade review

After a report, document the surprise, the gap and the reaction size for trades reviewed around the event.

ora-fin-0.1.1JPM5 questions

StateJPM
Questions5 asked
Answersreal, captured
Actionsnone

State

{"symbol":"JPM"} The server fills in the latest completed session.

Questions

`recent`	rule	`days_since_earnings <= 5` Reported in the last 5 days
`surprise`	rule	`eps_surprise_pct > 5` EPS beat by more than 5%
`beat`	English → rule	Did the stock beat on both EPS and revenue last quarter?
`gap`	English → rule	Did the stock gap up after its last earnings?
`react`	rule	`abs(last_earnings_move) > 5` Last earnings move bigger than 5%

Answers

Question	Answer	Reading
`recent` Reported in the last 5 days	0.0537	false
`surprise` EPS beat by more than 5%	0.0844	false
`beat` Did the stock beat on both EPS and revenue last quarter compiled to `beat_eps_last > 0.5 and beat_revenue_last > 0.5`	0.1409	false
`gap` Did the stock gap up after its last earnings compiled to `last_earnings_gap > 0`	0.9463	true
`react` Last earnings move bigger than 5%	0.0664	false

What it means

For JPM on the session of 2026-10-30: True now: Did the stock gap up after its last earnings. Not true: Reported in the last 5 days; EPS beat by more than 5%; Did the stock beat on both EPS and revenue last quarter; Last earnings move bigger than 5%.

These are the public facts of the event; reviewers compare them to the timing of trades.

Request JSON

{
  "model": "ora-fin-0.1.1",
  "state": {
    "symbol": "JPM"
  },
  "questions": {
    "recent": {
      "type": "yes_no",
      "ask": "days_since_earnings <= 5"
    },
    "surprise": {
      "type": "yes_no",
      "ask": "eps_surprise_pct > 5"
    },
    "beat": {
      "type": "yes_no",
      "instructions": "Did the stock beat on both EPS and revenue last quarter?"
    },
    "gap": {
      "type": "yes_no",
      "instructions": "Did the stock gap up after its last earnings?"
    },
    "react": {
      "type": "yes_no",
      "ask": "abs(last_earnings_move) > 5"
    }
  }
}

Response JSON (JPM, real)

{
  "model": "ora-fin-0.1.1",
  "answers": {
    "recent": {
      "yes_no": 0.0537,
      "rule": "days_since_earnings <= 5",
      "basis": "rule"
    },
    "surprise": {
      "yes_no": 0.0844,
      "rule": "eps_surprise_pct > 5",
      "basis": "rule"
    },
    "beat": {
      "yes_no": 0.1409,
      "rule": "beat_eps_last > 0.5 and beat_revenue_last > 0.5",
      "basis": "pattern"
    },
    "gap": {
      "yes_no": 0.9463,
      "rule": "last_earnings_gap > 0",
      "basis": "pattern"
    },
    "react": {
      "yes_no": 0.0664,
      "rule": "abs(last_earnings_move) > 5",
      "basis": "rule"
    }
  },
  "context": {
    "symbol": "JPM",
    "as_of": "2026-10-30"
  }
}

abs beat_eps_last beat_revenue_last days_since_earnings eps_surprise_pct last_earnings_gap last_earnings_move

Try it in the Playground

Compliance officer · example 5

Dividend-capture pattern review: KO, PG, T, VZ ex-dates

Flag accounts trading around ex-dates by first knowing which names went ex recently.

ora-fin-0.1.14 symbols, one request each4 questions

State4 symbols
Questions4 asked
Answersreal, captured
Actionsnone

State

{"symbol":"KO"} then PG, T, VZ, one request each. The server fills in the latest completed session.

Questions

`ex_recent`	rule	`days_since_ex_div <= 5` Went ex-dividend in the last five days
`ex_soon`	rule	`days_to_ex_div <= 30` Goes ex-dividend within 30 days
`yield3`	rule	`div_yield_ttm > 3` Yield above 3%
`pays`	rule	`pays_dividend > 0.5` Pays a dividend

Answers

Symbol	`ex_recent`rule	`ex_soon`rule	`yield3`rule	`pays`rule
`KO`	0.0537	0.2049	0.0537	0.8591
`PG`	0.2049	0.0537	0.2091	0.8591
`T`	0.0537	0.0537	0.9463	0.8591
`VZ`	0.0537	0.0537	0.9463	0.8591

true false not known for rules; forecasts are odds. Hover a column for its question.

What it means

Session of 2026-10-30, 4 symbols: Went ex-dividend in the last five days: true for none of 4. Goes ex-dividend within 30 days: true for none of 4. Yield above 3%: true for T, VZ of 4. Pays a dividend: true for KO, PG, T, VZ of 4.

Knowing the ex-dates is the first step; the trade review itself happens in the firm's own systems.

Request JSON (KO; the others differ only in the symbol)

{
  "model": "ora-fin-0.1.1",
  "state": {
    "symbol": "KO"
  },
  "questions": {
    "ex_recent": {
      "type": "yes_no",
      "ask": "days_since_ex_div <= 5"
    },
    "ex_soon": {
      "type": "yes_no",
      "ask": "days_to_ex_div <= 30"
    },
    "yield3": {
      "type": "yes_no",
      "ask": "div_yield_ttm > 3"
    },
    "pays": {
      "type": "yes_no",
      "ask": "pays_dividend > 0.5"
    }
  }
}

Response JSON (KO, real)

{
  "model": "ora-fin-0.1.1",
  "answers": {
    "ex_recent": {
      "yes_no": 0.0537,
      "rule": "days_since_ex_div <= 5",
      "basis": "rule"
    },
    "ex_soon": {
      "yes_no": 0.2049,
      "rule": "days_to_ex_div <= 30",
      "basis": "rule"
    },
    "yield3": {
      "yes_no": 0.0537,
      "rule": "div_yield_ttm > 3",
      "basis": "rule"
    },
    "pays": {
      "yes_no": 0.8591,
      "rule": "pays_dividend > 0.5",
      "basis": "rule"
    }
  },
  "context": {
    "symbol": "KO",
    "as_of": "2026-10-30"
  }
}

days_since_ex_div days_to_ex_div div_yield_ttm pays_dividend

Try it in the Playground

Compliance officer · example 6

Corporate actions to reconcile: recent and announced splits

Make sure position and limit records are adjusted for splits in names the firm trades.

ora-fin-0.1.13 symbols, one request each4 questions1 action

State3 symbols
Questions4 asked
Answersreal, captured
Actions1 rule · dry run

State

{"symbol":"NVDA"} then SMCI, AAPL, one request each. The server fills in the latest completed session.

Questions

`recent`	rule	`days_since_split <= 30` Split in the last 30 days
`announced`	rule	`days_to_split <= 365` A split is announced
`reverse`	English → rule	Has the stock done a reverse split in the past year?
`ratio`	English → rule	Was the last split at least 2-for-1?

Answers

Symbol	`recent`rule	`announced`rule	`reverse`rule	`ratio`rule	Fired
`NVDA`	0.0537	0.0537	0.0537	0.9463	none
`SMCI`	0.0537	0.0537	0.0537	0.9463	none
`AAPL`	0.0537	0.0537	0.0537	0.9463	none

true false not known for rules; forecasts are odds. Hover a column for its question.

Actions

Rules on the answers decide what to send. Each symbol runs its own request, so each one can fire.

recent >= 0.6 or announced >= 0.6→Emailcompliance-ops

NVDA 0 of 1 fired · dry run, nothing sent

SMCI 0 of 1 fired · dry run, nothing sent

AAPL 0 of 1 fired · dry run, nothing sent

What it means

Session of 2026-10-30, 3 symbols: Split in the last 30 days: true for none of 3. A split is announced: true for none of 3. Has the stock done a reverse split in the past year: true for none of 3. Was the last split at least 2-for-1: true for NVDA, SMCI, AAPL of 3. reconcile fired for none.

A split changes share counts in every record; the email is a reconciliation reminder, not an exception.

Request JSON (NVDA; the others differ only in the symbol)

{
  "model": "ora-fin-0.1.1",
  "state": {
    "symbol": "NVDA"
  },
  "questions": {
    "recent": {
      "type": "yes_no",
      "ask": "days_since_split <= 30"
    },
    "announced": {
      "type": "yes_no",
      "ask": "days_to_split <= 365"
    },
    "reverse": {
      "type": "yes_no",
      "instructions": "Has the stock done a reverse split in the past year?"
    },
    "ratio": {
      "type": "yes_no",
      "instructions": "Was the last split at least 2-for-1?"
    }
  },
  "actions": [
    {
      "id": "reconcile",
      "when": "recent >= 0.6 or announced >= 0.6",
      "do": "email",
      "connection": "compliance-ops",
      "subject": "Split reconciliation: {{symbol}}",
      "message": "{{symbol}} ({{as_of}}): split in the last 30 days {{recent|pct}} or announced {{announced|pct}}. Confirm positions, limits and restricted-list quantities are adjusted."
    }
  ],
  "dry_run": true
}

Response JSON (NVDA, real)

{
  "model": "ora-fin-0.1.1",
  "answers": {
    "recent": {
      "yes_no": 0.0537,
      "rule": "days_since_split <= 30",
      "basis": "rule"
    },
    "announced": {
      "yes_no": 0.0537,
      "rule": "days_to_split <= 365",
      "basis": "rule"
    },
    "reverse": {
      "yes_no": 0.0537,
      "rule": "days_since_reverse_split <= 365",
      "basis": "pattern"
    },
    "ratio": {
      "yes_no": 0.9463,
      "rule": "last_split_ratio >= 2",
      "basis": "pattern"
    }
  },
  "context": {
    "symbol": "NVDA",
    "as_of": "2026-10-30"
  },
  "actions": [
    {
      "id": "reconcile",
      "fired": false,
      "status": "skipped"
    }
  ]
}

days_since_reverse_split days_since_split days_to_split last_split_ratio

Try it in the Playground

Compliance officer · example 7

Close-location context for a marking-the-close review

When reviewing end-of-day activity, record whether the close sat at an extreme of the day's range and on what volume.

ora-fin-0.1.1SPY5 questions

StateSPY
Questions5 asked
Answersreal, captured
Actionsnone

State

{"symbol":"SPY"} The server fills in the latest completed session.

Questions

`top`	rule	`(C - L) / (H - L) > 0.75` Closed in the top quarter of the day's range
`bottom`	rule	`(C - L) / (H - L) < 0.25` Closed in the bottom quarter of the day's range
`vol`	rule	`V > avg(V, 20)` Volume above its 20-day average
`trend_day`	rule	`(O <= L + 0.2 * (H - L)) and (C >= H - 0.2 * (H - L))` A trend day up: opened near the low, closed near the high
`marubozu`	rule	`marubozu()` marubozu candle: Body at least 90% of the range: almost no wicks.

Answers

Question	Answer	Reading
`top` Closed in the top quarter of the day's range	0.1294	false
`bottom` Closed in the bottom quarter of the day's range	0.1492	false
`vol` Volume above its 20-day average	0.9463	true
`trend_day` A trend day up: opened near the low, closed near the high	0.0537	false
`marubozu` marubozu candle: Body at least 90% of the range: almost no wicks.	0.0537	false

What it means

For SPY on the session of 2026-10-30: True now: Volume above its 20-day average. Not true: Closed in the top quarter of the day's range; Closed in the bottom quarter of the day's range; A trend day up: opened near the low, closed near the high; marubozu candle: Body at least 90% of the range: almost no wicks..

A close at an extreme on a trend day is normal market behaviour; the context helps reviewers separate that from unusual activity.

Request JSON

{
  "model": "ora-fin-0.1.1",
  "state": {
    "symbol": "SPY"
  },
  "questions": {
    "top": {
      "type": "yes_no",
      "ask": "(C - L) / (H - L) > 0.75"
    },
    "bottom": {
      "type": "yes_no",
      "ask": "(C - L) / (H - L) < 0.25"
    },
    "vol": {
      "type": "yes_no",
      "ask": "V > avg(V, 20)"
    },
    "trend_day": {
      "type": "yes_no",
      "ask": "(O <= L + 0.2 * (H - L)) and (C >= H - 0.2 * (H - L))"
    },
    "marubozu": {
      "type": "yes_no",
      "ask": "marubozu()"
    }
  }
}

Response JSON (SPY, real)

{
  "model": "ora-fin-0.1.1",
  "answers": {
    "top": {
      "yes_no": 0.1294,
      "rule": "(C - L) / (H - L) > 0.75",
      "basis": "rule"
    },
    "bottom": {
      "yes_no": 0.1492,
      "rule": "(C - L) / (H - L) < 0.25",
      "basis": "rule"
    },
    "vol": {
      "yes_no": 0.9463,
      "rule": "V > avg(V, 20)",
      "basis": "rule"
    },
    "trend_day": {
      "yes_no": 0.0537,
      "rule": "(O <= L + 0.2 * (H - L)) and (C >= H - 0.2 * (H - L))",
      "basis": "rule"
    },
    "marubozu": {
      "yes_no": 0.0537,
      "rule": "marubozu()",
      "basis": "rule"
    }
  },
  "context": {
    "symbol": "SPY",
    "as_of": "2026-10-30"
  }
}

avg marubozu

Try it in the Playground

Compliance officer · example 8

Daily shock-day scan across traded large caps

Scan the firm's most-traded names for return shocks and record each hit in the surveillance system.

ora-fin-0.1.15 symbols, one request each4 questions1 actionworkflow · Mon–Fri at 17:30 ET

ScheduleMon–Fri at 17:30 ET
State5 symbols
Questions4 asked
Answersreal, captured
Actions1 rule · dry run

State

{"symbol":"AAPL"} then MSFT, AMZN, TSLA, NFLX, one request each. The server fills in the latest completed session.

Questions

`shock`	rule	`zscore(C / C[-1] - 1, 20) < -2` A two-sigma down day
`zs`	rule	`zscore(C / C[-1] - 1, 60) > 2` Daily return over two standard deviations above its 60-day norm
`vol`	rule	`V > 2 * ref(avg(V, 20), -1)` Volume more than twice the previous 20-day average
`outside`	rule	`outside_bar()` outside bar candle: Today's high above yesterday's and low below it.

Answers

Symbol	`shock`rule	`zs`rule	`vol`rule	`outside`rule	Fired
`AAPL`	0.253	0.0537	0.2473	0.2273	none
`MSFT`	0.1707	0.0537	0.262	0.0651	none
`AMZN`	0.089	0.0537	0.9135	0.0661	none
`TSLA`	0.0945	0.0537	0.0589	0.3172	none
`NFLX`	0.0639	0.0723	0.0708	0.401	none

true false not known for rules; forecasts are odds. Hover a column for its question.

Actions

Rules on the answers decide what to send. Each symbol runs its own request, so each one can fire.

shock >= 0.6 or zs >= 0.6→Webhooksurveillance-cases

AAPL 0 of 1 fired · dry run, nothing sent

MSFT 0 of 1 fired · dry run, nothing sent

AMZN 0 of 1 fired · dry run, nothing sent

TSLA 0 of 1 fired · dry run, nothing sent

NFLX 0 of 1 fired · dry run, nothing sent

What it means

Session of 2026-10-30, 5 symbols: A two-sigma down day: true for none of 5. Daily return over two standard deviations above its 60-day norm: true for none of 5. Volume more than twice the previous 20-day average: true for AMZN of 5. outside bar candle: Today's high above yesterday's and low below it.: true for none of 5. record fired for none.

Shocks in either direction are recorded with their volume context so later reviews start from facts.

As a scheduled workflow (Mon–Fri at 17:30 ET)

Create it under Workflows in the console: it runs these questions for every symbol on the schedule and fires the actions per symbol. The answers above are what one run returned.

{
  "name": "Daily shock-day scan across traded large caps",
  "symbols": "AAPL MSFT AMZN TSLA NFLX",
  "days": [
    1,
    2,
    3,
    4,
    5
  ],
  "at_time": "17:30",
  "model": "ora-fin-0.1.1",
  "questions": {
    "shock": {
      "type": "yes_no",
      "ask": "zscore(C / C[-1] - 1, 20) < -2"
    },
    "zs": {
      "type": "yes_no",
      "ask": "zscore(C / C[-1] - 1, 60) > 2"
    },
    "vol": {
      "type": "yes_no",
      "ask": "V > 2 * ref(avg(V, 20), -1)"
    },
    "outside": {
      "type": "yes_no",
      "ask": "outside_bar()"
    }
  },
  "actions": [
    {
      "id": "record",
      "when": "shock >= 0.6 or zs >= 0.6",
      "do": "webhook",
      "connection": "surveillance-cases",
      "body": {
        "symbol": "{{symbol}}",
        "as_of": "{{as_of}}",
        "type": "price_shock_context",
        "down_shock": "{{shock}}",
        "up_shock": "{{zs}}",
        "volume_spike": "{{vol}}",
        "note": "Context only; no finding."
      }
    }
  ]
}

Request JSON (AAPL; the others differ only in the symbol)

{
  "model": "ora-fin-0.1.1",
  "state": {
    "symbol": "AAPL"
  },
  "questions": {
    "shock": {
      "type": "yes_no",
      "ask": "zscore(C / C[-1] - 1, 20) < -2"
    },
    "zs": {
      "type": "yes_no",
      "ask": "zscore(C / C[-1] - 1, 60) > 2"
    },
    "vol": {
      "type": "yes_no",
      "ask": "V > 2 * ref(avg(V, 20), -1)"
    },
    "outside": {
      "type": "yes_no",
      "ask": "outside_bar()"
    }
  },
  "actions": [
    {
      "id": "record",
      "when": "shock >= 0.6 or zs >= 0.6",
      "do": "webhook",
      "connection": "surveillance-cases",
      "body": {
        "symbol": "{{symbol}}",
        "as_of": "{{as_of}}",
        "type": "price_shock_context",
        "down_shock": "{{shock}}",
        "up_shock": "{{zs}}",
        "volume_spike": "{{vol}}",
        "note": "Context only; no finding."
      }
    }
  ],
  "dry_run": true
}

Response JSON (AAPL, real)

{
  "model": "ora-fin-0.1.1",
  "answers": {
    "shock": {
      "yes_no": 0.253,
      "rule": "zscore(C / C[-1] - 1, 20) < -2",
      "basis": "rule"
    },
    "zs": {
      "yes_no": 0.0537,
      "rule": "zscore(C / C[-1] - 1, 60) > 2",
      "basis": "rule"
    },
    "vol": {
      "yes_no": 0.2473,
      "rule": "V > 2 * ref(avg(V, 20), -1)",
      "basis": "rule"
    },
    "outside": {
      "yes_no": 0.2273,
      "rule": "outside_bar()",
      "basis": "rule"
    }
  },
  "context": {
    "symbol": "AAPL",
    "as_of": "2026-10-30"
  },
  "actions": [
    {
      "id": "record",
      "fired": false,
      "status": "skipped"
    }
  ]
}

avg offset outside_bar ref zscore

Try it in the Playground

Compliance officer · example 9

AMD: verify chart claims before a research note publishes

Check the technical statements in a draft note against exact rules before supervisory approval.

ora-fin-0.1.1AMD5 questions1 action

StateAMD
Questions5 asked
Answersreal, captured
Actions1 rule · dry run

State

{"symbol":"AMD"} The server fills in the latest completed session.

Questions

`above200`	English → rule	Is the stock above its 200-day moving average?
`golden`	English → rule	Is the 50-day moving average above the 200-day?
`new_high`	English → rule	Did the stock make a new 52-week high today?
`rsi70`	English → rule	Is the stock overbought on RSI, above 70?
`macd`	English → rule	Is the MACD line above its signal line?

Answers

Question	Answer	Reading
`above200` Is the stock above its 200-day moving average compiled to `C > sma(C, 200)`	0.9284	true
`golden` Is the 50-day moving average above the 200-day compiled to `sma(C, 50) > sma(C, 200)`	0.9463	true
`new_high` Did the stock make a new 52-week high today compiled to `H > highest(H, 252)`	0.0537	false
`rsi70` Is the stock overbought on RSI, above 70 compiled to `rsi(C, 14) > 70`	0.055	false
`macd` Is the MACD line above its signal line compiled to `macd(C) > macdsignal(C)`	0.2076	false

Actions

Rules on the answers decide what to send.

above200 >= 0→Webhookreview-workflow

What each action would send 1 of 1 fired · dry run, nothing sent

ticketwould send (dry run)Webhook → review-workflow

{
  "symbol": "AMD",
  "as_of": "2026-10-30",
  "check": "technical_claims",
  "above_200d": 0.9284,
  "golden_setup": 0.9463,
  "new_52w_high": 0.0537,
  "rsi_over_70": 0.055,
  "macd_over_signal": 0.2076
}

What it means

For AMD on the session of 2026-10-30: True now: Is the stock above its 200-day moving average; Is the 50-day moving average above the 200-day. Not true: Did the stock make a new 52-week high today; Is the stock overbought on RSI, above 70; Is the MACD line above its signal line. Fired: ticket (Webhook → review-workflow).

Each claim in the draft is matched to a yes or no from the data; a claim the data does not support goes back to the author.

Request JSON

{
  "model": "ora-fin-0.1.1",
  "state": {
    "symbol": "AMD"
  },
  "questions": {
    "above200": {
      "type": "yes_no",
      "instructions": "Is the stock above its 200-day moving average?"
    },
    "golden": {
      "type": "yes_no",
      "instructions": "Is the 50-day moving average above the 200-day?"
    },
    "new_high": {
      "type": "yes_no",
      "instructions": "Did the stock make a new 52-week high today?"
    },
    "rsi70": {
      "type": "yes_no",
      "instructions": "Is the stock overbought on RSI, above 70?"
    },
    "macd": {
      "type": "yes_no",
      "instructions": "Is the MACD line above its signal line?"
    }
  },
  "actions": [
    {
      "id": "ticket",
      "when": "above200 >= 0",
      "do": "webhook",
      "connection": "review-workflow",
      "body": {
        "symbol": "{{symbol}}",
        "as_of": "{{as_of}}",
        "check": "technical_claims",
        "above_200d": "{{above200}}",
        "golden_setup": "{{golden}}",
        "new_52w_high": "{{new_high}}",
        "rsi_over_70": "{{rsi70}}",
        "macd_over_signal": "{{macd}}"
      }
    }
  ],
  "dry_run": true
}

Response JSON (AMD, real)

{
  "model": "ora-fin-0.1.1",
  "answers": {
    "above200": {
      "yes_no": 0.9284,
      "rule": "C > sma(C, 200)",
      "basis": "comparison"
    },
    "golden": {
      "yes_no": 0.9463,
      "rule": "sma(C, 50) > sma(C, 200)",
      "basis": "comparison"
    },
    "new_high": {
      "yes_no": 0.0537,
      "rule": "H > highest(H, 252)",
      "basis": "pattern"
    },
    "rsi70": {
      "yes_no": 0.055,
      "rule": "rsi(C, 14) > 70",
      "basis": "pattern"
    },
    "macd": {
      "yes_no": 0.2076,
      "rule": "macd(C) > macdsignal(C)",
      "basis": "comparison"
    }
  },
  "context": {
    "symbol": "AMD",
    "as_of": "2026-10-30"
  },
  "actions": [
    {
      "id": "ticket",
      "fired": true,
      "status": "dry_run",
      "do": "webhook",
      "connection": "review-workflow",
      "preview": {
        "body": {
          "symbol": "AMD",
          "as_of": "2026-10-30",
          "check": "technical_claims",
          "above_200d": 0.9284,
          "golden_setup": 0.9463,
          "new_52w_high": 0.0537,
          "rsi_over_70": 0.055,
          "macd_over_signal": 0.2076
        }
      }
    }
  ]
}

highest macd macdsignal rsi sma

Try it in the Playground

Compliance officer · example 10

QQQ marketing copy: is 'near all-time highs' accurate today?

Before a client communication goes out, verify market statements in it are true as of the latest session.

ora-fin-0.1.1QQQ4 questions

StateQQQ
Questions4 asked
Answersreal, captured
Actionsnone

State

{"symbol":"QQQ"} The server fills in the latest completed session.

Questions

`near_high`	rule	`C >= 0.98 * max(H, 252)` Within 2% of the 52-week high
`new_high`	rule	`H > highest(H, 252)` New 52-week high
`up_month`	English → rule	Is the stock up over the last month?
`up_3m`	English → rule	Is the stock up more than 10% over the last three months?

Answers

Question	Answer	Reading
`near_high` Within 2% of the 52-week high	0.0537	false
`new_high` New 52-week high	0.0537	false
`up_month` Is the stock up over the last month compiled to `C > C[-21]`	0.0664	false
`up_3m` Is the stock up more than 10% over the last three months compiled to `C / C[-63] - 1 > 0.1`	0.0537	false

What it means

For QQQ on the session of 2026-10-30: Not true: Within 2% of the 52-week high; New 52-week high; Is the stock up over the last month; Is the stock up more than 10% over the last three months.

Marketing statements must be accurate when sent; if a rule is false the copy is changed, not the rule.

Request JSON

{
  "model": "ora-fin-0.1.1",
  "state": {
    "symbol": "QQQ"
  },
  "questions": {
    "near_high": {
      "type": "yes_no",
      "ask": "C >= 0.98 * max(H, 252)"
    },
    "new_high": {
      "type": "yes_no",
      "ask": "H > highest(H, 252)"
    },
    "up_month": {
      "type": "yes_no",
      "instructions": "Is the stock up over the last month?"
    },
    "up_3m": {
      "type": "yes_no",
      "instructions": "Is the stock up more than 10% over the last three months?"
    }
  }
}

Response JSON (QQQ, real)

{
  "model": "ora-fin-0.1.1",
  "answers": {
    "near_high": {
      "yes_no": 0.0537,
      "rule": "C >= 0.98 * max(H, 252)",
      "basis": "rule"
    },
    "new_high": {
      "yes_no": 0.0537,
      "rule": "H > highest(H, 252)",
      "basis": "rule"
    },
    "up_month": {
      "yes_no": 0.0664,
      "rule": "C > C[-21]",
      "basis": "pattern"
    },
    "up_3m": {
      "yes_no": 0.0537,
      "rule": "C / C[-63] - 1 > 0.1",
      "basis": "pattern"
    }
  },
  "context": {
    "symbol": "QQQ",
    "as_of": "2026-10-30"
  }
}

highest max offset

Try it in the Playground

Information, not investment advice. Rules answer from the latest completed daily session; forecasts are probabilities. Actions ran as dry runs: nothing was sent. More in Sell side: Equity research analyst, Sell-side sales trader, Execution trader / desk, Market maker, Prop trading firm.

Create an account Open the Playground Everything here also lives in the console, next to your keys and usage.

Individuals

Buy side

Sell side

Corporates

Builders

Institutions

NVDA: document unusual volume ahead of an earnings report

State

Questions

Answers

Actions

What it means

Restricted-list names: large moves to note in the daily log

State

Questions

Answers

Actions

What it means

TSLA gap documentation for the surveillance file

State

Questions

Answers

What it means

JPM earnings reaction: context for the trade review

State

Questions

Answers

What it means

Dividend-capture pattern review: KO, PG, T, VZ ex-dates

State

Questions

Answers

What it means

Corporate actions to reconcile: recent and announced splits

State

Questions

Answers

Actions

What it means

Close-location context for a marking-the-close review

State

Questions

Answers

What it means

Daily shock-day scan across traded large caps

State

Questions

Answers

Actions

What it means

AMD: verify chart claims before a research note publishes

State

Questions

Answers

Actions

What it means

QQQ marketing copy: is 'near all-time highs' accurate today?

State

Questions

Answers

What it means