Transit Range Scanner
The pack's future-teller for weekly and yearly horoscopes
- transit_data_json
Transit Data Calculator tells you the sky on one day. Transit Range Scanner is what you use when you want the whole month at once. It takes a start date and a duration, scans day by day, and reports the astrologically interesting events in the window: when each planet changes sign ("ingresses") and when the major moon phases land. The README calls it the "temporal scanner to predict the future," and it's the node the pack was really built around - feed its JSON to an LLM and you've automated the weekly or annual horoscope.
How it works
The setup is the familiar pack machinery: geocode city_name via Nominatim, resolve the timezone, then compute kerykeion subjects with local Swiss Ephemeris math. The scan itself is brute-force and honest about it:
- It anchors on your start date at noon (hour 12:00) and establishes a baseline by computing the chart for the day before the window.
- For each day in the range, it computes a fresh
AstrologicalSubjectand compares it to the previous day's. - Two kinds of events get flagged:
- Ingresos - a planet crossing into a new sign, e.g.
"Ingreso: Mercurio entra en Piscis". - Major moon phases - when the day's phase flips to New Moon, Full Moon, First Quarter, or Last Quarter, reported as e.g.
"Luna Llena en Leo".
- Ingresos - a planet crossing into a new sign, e.g.
The output JSON also includes a clima_astral_general summary of the Sun and Mars signs on the start date, which gives an LLM the "general vibe" of the period to write around.
Because each day means a full chart computation, a full-length scan is doing ~367 ephemeris calculations for a 366-day run. That's not slow in absolute terms - Swiss Ephemeris is fast - but it's the heaviest node in the pack, so don't expect instant results on a year-long sweep.
The inputs that matter
- start_year, start_month, start_day - where the window begins. Years limited to 1900–2100.
- duration_days - 1 to 366. 7 gives you a week, 30-ish a month, 365 a full year. This is the dial that decides whether you get a horoscope or an almanac.
- city_name - the reference location for timezone and moon-phase boundaries. Live geocoder lookup, so it needs internet.
The single output
- transit_data_json - a STRING with
periodo(the date range, asDel 01/01/2024 al 07/01/2024 (7 días)),eventos_clave(the list of ingress and moon-phase events, each with afechaand a Spanish description), andclima_astral_general.
This is the pack's flagship text-workflow: connect the scanner's output to an LLM node (Ollama, or whatever you run locally) with a system prompt like "You are a fortune teller. With this natal chart and these week's transits, write a prediction." Combine it with a Natal Chart Calculator for a genuinely structured "your chart meets this week's sky" reading - the README's featured example workflow, and it's a good one.
Installing it
Same pack install as the rest:
cd ComfyUI/custom_nodes
git clone https://github.com/Meisoftcoltd/Comfyui-Kerykeion
cd Comfyui-Kerykeion
pip install -r requirements.txt
or "ComfyUI-Kerykeion" via ComfyUI Manager and restart. No model downloads - kerykeion, geopy, timezonefinder, cairosvg, Pillow all arrive via pip, and the astronomy runs locally.
Common issues
{"error": "Ciudad no encontrada..."}- the one geocoding call failed (or you're offline). Same failure mode as the other calculators; check the console for theWarning (TransitRangeScanner)line.- Long scans feel slow. 366 days means ~367 chart computations. A week is instant; a year takes a noticeable beat. That's inherent to the brute-force approach, not a hang.
- The noon anchor. Every day is evaluated at 12:00, so an ingress that happens at 3 a.m. is logged against that day - accurate enough for horoscope writing, not for down-to-the-hour astronomy.
- Spanish output.
Ingreso,Luna Llena,Cuarto Creciente- it's the pack's language, and your LLM should be fine with it.
If your goal is automated weekly or yearly readings, this is the node worth learning first. It's the one that turns a ComfyUI graph into something that writes the horoscopes instead of reading them.
Inputs (5)
| Name | Type | Default | Description |
|---|---|---|---|
| start_year | INT | 20241900–2100 | — |
| start_month | INT | 11–12 | — |
| start_day | INT | 11–31 | — |
| duration_days | INT | 71–366 | — |
| city_name | STRING | Madrid | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| transit_data_json | STRING | — |