VoiceFilterChain::applyHighPass collapsed to a constant-gain multiplier instead of a real high-pass response. The inner loop read `samples[prevIdx]` as "previous input" — but that slot held y[n-1] (output) since the previous iteration had overwritten it. Substituting that into the formula y[n] = α(y[n-1] + x[n] - x[n-1]) with x[n-1] := y[n-1] gave y[n] = α(y[n-1] + x[n] - y[n-1]) = α · x[n] i.e. just a fixed gain ≈ 0.81 at 300Hz cutoff / 8kHz. DC offsets sailed through; the chain only kept signal levels reasonable because the AGC stage downstream pulled the residual toward target. The same bug exists in upstream LXST-kt's native_audio_filters.cpp (filed as LXST-kt#13). Fix: walk per-channel with explicit `xPrev` / `yPrev` variables so input history isn't clobbered by the output write. Per-chunk save of `lastInputs[ch]` now stores the actual last input, not the last output, so the first sample of each new chunk uses the correct x[n-1]. Test `dc_offset_attenuated_by_hpf` updated to assert tail RMS < 0.01 (was < 0.5 — accommodating the broken behavior). Pass. End-to-end acoustic test (Mac speaker → T-Deck mic): pyxis_rms went up from 4378 → 5835, bot_rms from 1194 → 1517 — DC offset removal lets clean signal through better. Pyxis-side fix only. LXST-kt should pick up the same fix upstream — see project_lxst_hpf_filter_bug.md in the vault. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Pyxis Tests
Three test surfaces, each runnable independently.
1. Pyxis-unique pytest suite
Native C++ tests of pyxis-unique code (BLE fragmenter, HDLC, etc.) and Python tests of build scripts.
/usr/bin/python3 -m pytest tests/build_scripts tests/native -v
System Python 3.9 has pytest pre-installed; Homebrew Python does not.
build_scripts/test_patch_nimble.py— verifiespatch_nimble.pyidempotency, drift detection, missing-file handlingnative/test_hdlc.{cpp,py}— HDLC escape/unescape/frame round-trip + golden vector against Python RNSnative/test_ble_fragmenter.{cpp,py}— BLEFragmenter ↔ BLEReassembler: in-order, out-of-order, duplicate, dropped+timeout, per-peer isolation, MTU changenative/test_ble_peer_manager.{cpp,py}— connection-map state machine: discover, identity promotion, blacklist, handle map cleanup, MAC rotation, pool exhaustionnative/test_ble_operation_queue.{cpp,py}— GATT op queue: FIFO, busy-state, timeout, clearForConnection, buildernative/test_ring_buffers.{cpp,py}— PCM + encoded SPSC ring buffers, including 100k-frame multithreaded producer/consumer stressnative/test_audio_filters.{cpp,py}— VoiceFilterChain frequency response, peak limiting, multichannel
Adding a new native C++ test
Pattern: pure C++ unit + thin pytest wrapper.
- Write
tests/native/test_<thing>.cpp— include from../../{src,lib/...}for the unit under test, use the existingEXPECT_EQ/EXPECT_TRUE/RUN(name)framework, return non-zero on any failure. - If the unit pulls in microReticulum types, the shims in
tests/native/cover what's been needed so far:Bytes.h→bytes_shim.h(minimalRNS::Bytes— append/data/size/writable/resize/mid)Log.h(no-opTRACE/WARNING/INFO/ERRORmacros)Utilities/OS.h(OS::time()withset_fake_time()/clear_fake_time())
- Write
tests/native/test_<thing>.py— copy thetest_hdlc.pytemplate, swap source/include paths, run. - Confirm:
/usr/bin/python3 -m pytest tests/native/test_<thing>.py -v
2. LXST audio interop tests
Python tests verifying wire format and codec compatibility between pyxis (C++), Python LXST, and LXST-kt (Kotlin).
/usr/bin/python3 -m pytest tests/interop -v
Requires pycodec2 and access to $HOME/repos/public/LXST.
3. microReticulum native unit tests (PlatformIO)
The fork's microReticulum tests live under deps/microReticulum/test/ and run via PlatformIO Unity.
cd deps/microReticulum && pio test -e native17
Use native17, not native — the C++11 env is broken (static-constexpr ODR-use). Baseline as of 2026-05-02: 94/114 PASS, 7 FAIL, 5 SKIPPED, 7 suites ERRORED. The clean suites are
test_os, test_bytes, test_msgpack, test_crypto, test_filesystem, test_objects, test_interop, test_general, test_reference, test_example, test_collections.