Feature: Importance-based event classification #22

New issue

Open

opened 2025-12-11 00:06:46 +00:00 by blightbow · 0 comments

blightbow commented

2025-12-11 00:06:46 +00:00

Owner

Overview

Update message classification to return PerceptionEvent with importance scoring rather than routing to separate buffers.

Classification Changes

Current: classify_message() returns ("TRIGGER" | "CONTEXT" | "IGNORE", metadata)

New: classify_message() returns PerceptionEvent with attributes:

requires_response: True for TRIGGER conditions
importance: Scored based on multiple factors
All events go to unified perception_stream

Importance Scoring Heuristics

Condition	Importance
Direct @mention of assistant	10
Direct message (page, whisper)	9
Speech from permitted sender	8
Pose/action in same room	6
Channel message	5
Other assistant's output	4
System notification	3

Files to Modify

messaging/classification.py - Return PerceptionEvent, add importance scoring
event_handling.py - New record_perception_event(), record_self_action()
tool_execution.py - Call record_self_action() after tool execution

Phase

Phase 2 of Milestone #4

## Overview Update message classification to return `PerceptionEvent` with importance scoring rather than routing to separate buffers. ## Classification Changes Current: `classify_message()` returns `("TRIGGER" | "CONTEXT" | "IGNORE", metadata)` New: `classify_message()` returns `PerceptionEvent` with attributes: - `requires_response`: True for TRIGGER conditions - `importance`: Scored based on multiple factors - All events go to unified `perception_stream` ## Importance Scoring Heuristics | Condition | Importance | |-----------|------------| | Direct @mention of assistant | 10 | | Direct message (page, whisper) | 9 | | Speech from permitted sender | 8 | | Pose/action in same room | 6 | | Channel message | 5 | | Other assistant's output | 4 | | System notification | 3 | ## Files to Modify - `messaging/classification.py` - Return PerceptionEvent, add importance scoring - `event_handling.py` - New `record_perception_event()`, `record_self_action()` - `tool_execution.py` - Call `record_self_action()` after tool execution ## Phase Phase 2 of Milestone #4