personal memory agent
Show full context keys in tokens app instead of truncating
Remove the two-part prefix truncation in _parse_context_prefix() so the
tokens app displays granular context keys like observe.describe.frame and
observe.describe.meeting instead of aggregating them all under observe.describe.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
+2
-5
+2
-5
apps/tokens/routes.py
+2
-5
apps/tokens/routes.py
···
30
30
31
31
32
32
def _parse_context_prefix(context: str) -> str:
33
-
"""Extract first two parts of context (e.g., 'agent.default.123' -> 'agent.default')."""
33
+
"""Return full context string for grouping."""
34
34
if not context:
35
35
return "unknown"
36
-
parts = context.split(".")
37
-
if len(parts) >= 2:
38
-
return f"{parts[0]}.{parts[1]}"
39
-
return parts[0] if parts else "unknown"
36
+
return context
40
37
41
38
42
39
def _aggregate_token_data(day: str) -> Dict[str, Any]: