Test Suite 3 — Boolean mode

Harness function	suite3() in test-scripts/test-harness.js
Checks	15 (the # column below reconciles)
Kind	Synchronous

Proves boolean coercion (default and custom value lists), the fraction-based vote rules and their latch edges, the minority-report change-only semantics, and command precedence over coercion.

Baseline config (applies unless a scenario says otherwise): boolean node — Majority trigger rule, Same as trigger release rule, default True/False value lists, quorum and staleness disabled, Emit aggregate on change only checked, heartbeat off, persistence off.

Traceability convention: the Check column is the harness's literal PASS/FAIL string — a failing line in the test output can be searched in this document verbatim. This document moves with the harness: a new, renamed, or removed check lands here in the same change.

---

3.1 Default coercion lists

Five sources fed true, "ON", "yes", "1", 1; then three of them flipped with "OFF", "no", 0; then an unlisted string.

#	Check	Protects
1	default true-list coerces true/ON/yes/'1'/1 (fraction 1.0)	Native booleans/1 and the default true-list all coerce, case-insensitively
2	default false-list coerces OFF/no/0 (fraction 2/5)	The default false-list coerces the same way, and the aggregate is the fraction of fresh sources true
3	unlisted value rejected	Anything matching neither list is a rejected reading (ignored:true) — never silently skews the vote

3.2 Custom lists replace the defaults

Config: True Values open, False Values closed.

#	Check	Protects
4	custom true value 'OPEN' accepted and triggers (any/majority of 1)	Custom vocabularies (door contacts etc.) coerce case-insensitively
5	default 'true' rejected when custom lists replace defaults	Custom lists replace the defaults rather than extending them

3.3 Majority latch behavior and minority reports

Sources d1/d2/d3 walked through true, exact split, and minority states.

#	Check	Protects
6	1/1 majority triggers	The denominator is the fresh source count — one sensor can satisfy Majority
7	exact 0.5 stays latched (majority >= 0.5)	The Majority boundary is ≥ 0.5: an exact split holds the latch
8	exact split produces no minority report	No consensus exists at an exact split, so no source is a "minority"
9	1/3 releases under sameastrigger	Same as trigger releases the moment the trigger rule stops being satisfied
10	minority report fires with [d1]	A fresh source disagreeing with the consensus is named in msg.minorityTopics
11	unchanged minority set does not re-report	Reports fire on set change only — a stuck sensor produces one report, not a stream
12	minority resolution (set -> empty) reports once	Resolution is itself a change: one final report with an empty set

3.4 At-least-N configured preset

Config: trigger rule At least N with N = 2.

#	Check	Protects
13	atleastn(2): 1 true does not trigger	The N threshold counts true sources, not fractions
14	atleastn(2): 2 true triggers	…and fires exactly at N

3.5 Command precedence over coercion

A payload of "RESET" sent with a topic on a boolean node.

#	Check	Protects
15	command string wins over coercion even with a topic	A recognized command is always a command, in both type modes — command words can never function as boolean values