|
@globalcaos/openclaw-computational-humor
|
Your agent jokes. You laugh — it remembers. Humor from embedding geometry, calibrated by your reactions. |
Apache-2.0 |
1
greenflagged
/ 1 total
|
|
@globalcaos/openclaw-fractal-reflection
|
Your agent finishes the turn — then reflects on it across four levels: specific, pattern, system, worldview. |
Apache-2.0 |
1
greenflagged
/ 1 total
|
|
@globalcaos/openclaw-identity-persistence
|
All agents drift from their persona — we measure it with EWMA SyncScore, re-inject SOUL.md before it's gone. |
Apache-2.0 |
1
greenflagged
/ 1 total
|
|
@globalcaos/openclaw-learned-intuition
|
Five neural networks vote yes/no on every tool call — trained on real failures, rules fallback if models miss. |
Apache-2.0 |
1
greenflagged
/ 1 total
|
|
@globalcaos/openclaw-memory-enhancements
|
Same concept, twentieth time — why search again? O(1) hippocampus index layered on upstream memory-core. |
Apache-2.0 |
1
greenflagged
/ 1 total
|
|
@globalcaos/openclaw-round-table
|
Three models, one answer — cheaper than calling Opus once. RAAC debate protocol with cognitive diversity scoring. |
Apache-2.0 |
1
greenflagged
/ 1 total
|
|
@globalcaos/openclaw-total-recall
|
Yesterday's chat got compacted away? Per-session event store with FTS + vector recall — pulled back on demand. |
Apache-2.0 |
1
greenflagged
/ 1 total
|