The app shipped with certificate validation bypassed globally and on by
default — InsecureTrustDelegate trusted any cert from any host. That was
the evaluation's P1: anyone on the LAN could MITM call audio, transcripts,
and voiceprints.
The backend's Start9 cert already validates under normal system trust when
the StartOS Root CA is installed in the keychain (confirmed: URLSession
default validation returns 200 against the backend and its fallback), so the
bypass is unnecessary:
- skip-TLS now defaults to off
- when explicitly enabled, the bypass is scoped to the configured host via
InsecureTrustDelegate.allowsTrustOverride, never "trust any server"
- the host gate is pure and unit-tested (InsecureTrustDelegateTests)
Docs reconciled: AGENTS.md backend/TLS line and Current state.
The backend shipped dual-channel mode; wire the client to it. We already capture
mic (you) and system (others) separately, so send them as two files instead of the
mono mix — fixing the misattribution at the source.
- SparkControlClient: labelMergeDual(mic_file, system_file, self_name, self_vad);
multipart generalized to N files; shared POST/retry/decode extracted.
- SessionPackager.rebasedSelfVadData: chunk-local [{start,end}] for self_vad;
sliceAudio reused for both tracks.
- TranscriptPipeline.process: dual-channel chunking (slice mic+system, rebase
timeline + self_vad per chunk) when system audio is healthy; mono mixed-file
fallback (self folded into the timeline) otherwise.
- VisualCapture.finish: write the full visual_timeline.json (remote + self merged)
but return REMOTE (vision) segments only — self travels via the mic channel.
- TranscriptAssembler: rank mic_channel highest (the user's own track wins).
- VoiceprintStore: store the clean mic_channel self voiceprint.
- SessionController: pass mic/system URLs + remote timeline + channel self-spans +
self_name + systemHealthy; self_vad.json now reflects the channel-verified spans.
Validated END-TO-END against the live backend on the real misattributing session:
'Go Bitcoin' (remote) is now attributed to Unknown_0, NOT the user; the user's own
lines come back source=mic_channel; per-channel ASR recovered fuller remote text.
36/36 XCTest (4 new: self_vad rebase, mic_channel ranking + voiceprint storage).
Grant Gilliam