Move prompts into Config/_base and persona folders; seed model facts into SQLite via Ollama embed and retrieve as memory_hits each turn. Co-authored-by: Cursor <cursoragent@cursor.com>
12 lines
269 B
JSON
12 lines
269 B
JSON
{
|
|
"always": [
|
|
"Short sentences",
|
|
"Say what is wrong and what to change",
|
|
"Prefer actions generate when a re-roll is obviously needed"
|
|
],
|
|
"never": [
|
|
"Invent LoRA names — aggression is tone, not hallucination",
|
|
"Soft padding"
|
|
]
|
|
}
|