Minimal Signalling Channel

Claim Type: mechanism_hypothesis
Scope: Minimal signalling channel
Depends On: ARC-009, ARC-010
Status: candidate
Claim ID: MECH-014

Source: docs/processed/legacy_tree/architecture/language/minimal_signalling_channel.md

Elaborates Section 5 (Social Extension: Language) of REE_CORE.md.

Minimal Signalling Channel (Pre-language)

Before full language, REE benefits from a minimal signalling interface that can externalise:

harm/degradation alerts
intent/commitment markers
uncertainty/confidence markers
“stop/avoid” priors (trajectory warnings)

Interface sketch

Inputs: internal summaries (z_theta/z_delta slices, residue flags, confidence) Outputs: discrete or continuous signals that other agents can condition on

Design constraints

Signals are integrated via trust-weighting (reliability inference)
Signals cannot mask embodied harm channels in the receiving agent
Signals should be cheap relative to full other-model inference

Open Questions

None noted in preserved sources.

MECH-014

References / Source Fragments

docs/processed/legacy_tree/architecture/language/minimal_signalling_channel.md