hermes-agent/gateway
nnnet 5582b51a68 fix(gateway): stop poisoning the LLM prompt with STT-mode chatter
The STT-failure enrichment templates injected setup instructions —
"no STT provider is configured", "a direct message has already been
sent", and a "hermes-agent-setup" skill mention — into the LLM-visible
prompt. That text persists in conversation history, so after one STT
failure the model kept volunteering Whisper/Vosk setup advice on every
later voice turn, even after transcription started working (observed in
prod on gpt-5-nano). The gateway also fired a hardcoded English notice
via _stt_adapter.send(), producing a second, wrong-language reply that
TTS then spoke aloud.

- Neutralize all enrichment templates: success passes the transcript
  through as a plain quoted line; every failure branch emits a single
  [voice message could not be transcribed] marker.
- Move the operator-facing failure cause to logger.info so it stays
  diagnosable in container logs without leaking into the prompt.
- Remove the hardcoded English _stt_adapter.send() notice; the LLM now
  produces one coherent reply in the user's language.
- Update the gateway STT tests to assert the neutral contract.

Co-authored-by: Hermes Agent <noreply@nousresearch.com>
2026-06-30 04:23:03 -07:00
..
assets
…
builtin_hooks
…
platforms feat(whatsapp): opt-in forwarding of owner-typed messages in bot mode 2026-06-30 03:41:43 -07:00
relay refactor(relay): adopt scope_id wire key (guild_id → scope_id dual-read/write) (#55289) 2026-06-30 11:16:53 +10:00
__init__.py
…
authz_mixin.py
…
cgroup_cleanup.py
…
channel_directory.py
…
code_skew.py
…
config.py feat(gateway): per-platform typing_indicator toggle 2026-06-29 21:12:57 -07:00
dead_targets.py fix(gateway): skip confirmed-dead delivery targets (deleted groups, blocked bots) (#55115) 2026-06-29 13:23:29 -07:00
delivery.py fix(gateway): skip confirmed-dead delivery targets (deleted groups, blocked bots) (#55115) 2026-06-29 13:23:29 -07:00
display_config.py
…
drain_control.py feat(gateway): suppress home-channel shutdown broadcast on flagged drains (#54824) 2026-06-29 12:18:11 -07:00
hooks.py
…
kanban_watchers.py
…
memory_monitor.py
…
message_timestamps.py
…
mirror.py
…
pairing.py
…
platform_registry.py
…
response_filters.py
…
restart.py
…
rich_sent_store.py
…
run.py fix(gateway): stop poisoning the LLM prompt with STT-mode chatter 2026-06-30 04:23:03 -07:00
runtime_footer.py
…
scale_to_zero.py
…
session.py fix(gateway): stop per-turn agent-cache eviction from model + message_id signature churn 2026-06-30 04:22:41 -07:00
session_context.py
…
shutdown_forensics.py
…
slash_access.py
…
slash_commands.py feat(gateway): show per-category context breakdown in /usage (#55204) 2026-06-29 20:42:19 -07:00
status.py
…
sticker_cache.py
…
stream_consumer.py fix(gateway): add session staleness guard to stream consumer 2026-06-30 03:42:25 -07:00
stream_dispatch.py
…
stream_events.py
…
whatsapp_identity.py
…