Relationship intelligence

Relationship Intelligence

Choose the workflow

• For one person, use the natal workflow and scripts/analyze_natal.py.
• For two people, use the relationship workflow and scripts/analyze_relationship.py.

Core workflow

1. Collect both people's birth date, local birth time, and place. Prefer explicit UTC offsets and coordinates when available.
2. Treat an unknown or approximate time as a material limitation. Do not invent a time, coordinates, houses, angles, or event predictions.
3. Normalize the input into JSON accepted by the appropriate bundled script. If coordinates or UTC offset are omitted, disclose that the script will use Nominatim and timezone boundary data to resolve them.
4. Run the bundled script and use its structured evidence as the factual basis of the response.
5. Separate calculated observations from interpretation. Cite the aspect, orb, and chart type behind each important claim.
6. Lead with the strongest repeated themes, then tensions, then practical reflection prompts.
7. State assumptions and confidence. Present astrology as a reflective framework, not a scientific measurement or certain prediction.

Single-person input

Use one birth-data object:

{
  "date": "1992-02-26",
  "time": "07:00",
  "place": "Hangzhou, Zhejiang, China"
}

Run from the plugin root:

python3 skills/relationship-intelligence/scripts/analyze_natal.py input.json --pretty

Return planetary sign positions, the tightest natal aspects, assumptions, limitations, and confidence. Do not infer houses, ascendant, or midheaven because this version does not calculate them.

Relationship input

Use one object with person_a and person_b:

{
  "person_a": {
    "date": "1991-08-15",
    "time": "10:40",
    "utc_offset": "+08:00",
    "latitude": 30.2741,
    "longitude": 120.1551,
    "place": "Hangzhou"
  },
  "person_b": {
    "date": "1994-03-21",
    "time": "01:00",
    "utc_offset": "+08:00",
    "latitude": 28.2282,
    "longitude": 112.9388,
    "place": "Changsha"
  }
}

latitude, longitude, and utc_offset may be omitted when place is specific enough to geocode. Manual values always take precedence. Place-name lookup requires network access; historical UTC offsets are derived for the supplied local birth date and time.

Run from the plugin root:

python3 skills/relationship-intelligence/scripts/analyze_relationship.py input.json

Use --pretty for readable JSON. If pyswisseph is unavailable, report the dependency error instead of fabricating results.

Interpretation guardrails

• Avoid categorical labels such as soulmate, doomed, faithful, abusive, or narcissistic.
• Never infer private facts, consent, intent, or future behavior from a chart.
• Do not rank people or recommend ending/starting a relationship solely from astrology.
• Prefer calibrated language: “may highlight,” “is often read as,” and “could be explored.”
• For medical, legal, financial, abuse, or crisis concerns, prioritize appropriate real-world professional support.

Relationship output

Return:

• a concise summary;
• strongest supportive and challenging evidence with aspect/orb;
• repeated themes across chart techniques;
• assumptions and missing-data limitations;
• confidence level;
• optional reflection questions grounded in the evidence.

Treat synastry, midpoint_composite, davison, and marks as distinct techniques. Do not merge their evidence or describe the midpoint composite as a Davison chart. A Marks calculation returns two charts: the Davison midpoint between the relationship event and each person's birth event. Label them person_a and person_b; never treat one person's chart as the other's internal experience.