TTS audio container mismatch and Opus conversion patches

NousResearch/hermes-agent · cluster 36131 · agent/transports

Data as of 2026-10-22 12:15 UTC

Nothing open on GitHub. Fix merged in #73507. Check it covers your case.

open issues

open PRs

items grouped

Known?: Yes. 11 related items grouped here, 3 issues already closed.
Fix event: A fix was merged to main in PR #73507. Check it covers your case. No release data is tracked.
Tracker: The maintainers' tracker marks this partially resolved, deciding PR #73507. That is separate from the fix event.
Still open: Nothing open on GitHub.

Next step: read #73507, then check it covers your case.

Read #73507 Pick up a Lab task

Several TTS providers (Edge, Piper, MiniMax, xAI) were saving non-Opus audio under .ogg extensions, breaking Telegram voice bubbles and on-device transcription. Patches add magic-byte container detection and proper transcoding to Opus before delivery.

Open issues0

No open issues.

Open PRs0

No open pull requests.

Fixes on record

Merged to main. A stamp counts the issues a fix is linked to.

#73507 fix(gateway): central magic-byte container detection for all inbound audio

merged to main

files: contributors/emails/[redacted:email], contributors/emails/[redacted:email], gateway/platforms/base.py, gateway/platforms/signal.py, plugins/platforms/telegram/adapter.py, tests/gateway/test_document_cache.py, tests/gateway/test_extract_local_files.py, tests/gateway/test_platform_base.py, tests/gateway/test_signal.py, tests/gateway/test_telegram_documents.py, tests/tools/test_audio_container.py, tests/tools/test_send_message_tool.py, tests/tools/test_transcription_tools.py, tools/audio_container.py, tools/send_message_tool.py, tools/transcription_tools.py, tools/tts_tool.py

7 PRs closed without merging

From the close comments: duplicate 2 · merged via another change 1 · stale or abandoned 1 · superseded by another PR 1 · other 2

#36166 fix(gateway): sniff cached audio container type · merged via another change → 156edc5, a1582a7, #73507
#36179 fix(gateway): sniff audio magic bytes in cache_audio_from_bytes to fix wrong .ogg extension · superseded by another PR → #36166, #36176
#57069 fix(tts): convert Edge TTS MP3 to Opus when caller supplies .ogg path · duplicate → #20882, #57048
#57080 fix(tts): convert explicit Edge OGG output · stale or abandoned
#57190 fix(tts): transcode edge ogg output paths · other → #73072

Show 2 more

#58848 fix(tts): convert Piper WAV-in-OGG to real Opus on Telegram · other → #73072
#59196 fix(tts): convert Piper raw WAV to Opus before sending as .ogg · duplicate → #58848, #58845, #59092

Closed issues · 3

Closed#36131[BUG]Audio attachments cached with wrong .ogg extension (real format MP4/QuickTime+AAC) → on-device transcription fails
Closed#57048Edge TTS provider silently saves MP3 bytes under a .ogg extension when given an explicit .ogg output_path (breaks Telegram voice bubbles)
Closed#58845[Bug]: Piper TTS writes raw WAV to .ogg output path — Opus conversion skipped, Telegram shows 0-second voice bubble

How this cluster was grouped

Link evidence (partly structural links): some members are connected by structural evidence (shared changed files, or a closing/superseded reference); the rest are joined by an automated relevance judgement over their text.
Members outside that structural core may belong to a different underlying problem.
Pairwise link density 0.16: the share of item pairs that carry a link, counting a structural link in full and a relevance-only link at 0.36. 1.00 means every pair is linked.

Pick up a Lab task

Check whether a fix covers the open issues here, with your own agent. Bring back evidence for maintainers.

Open the Lab

At a glance

Subsystem: agent/transports
Merged PRs: #73507
Closed issues: 3
Attempts: 7 PRs closed without merging
Fix link checked: 28 Jul 2026

For your agent

The same facts, typed.

cluster/36131.json
cluster/36131.md