Christian Gick e81aa79396 fix: increase voice PDF context to 40k chars, fix language detection sanity ...

- Voice context per-document limit 10k→40k chars (was cutting off at page 6)
- Language detection: reject results >30 chars (LLM returning sentences)
- Voice.py: generalize "PDF" label to "Dokumente"

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

2026-02-23 12:40:13 +02:00

memory-service

feat: Replace JSON memory with pgvector semantic search (MAT-11)

2026-02-20 06:25:50 +02:00

.gitignore

feat: DM translation workflow for forwarded foreign messages

2026-02-19 08:56:49 +02:00

agent.py

feat(e2ee): Add HKDF E2EE support for Element Call compatibility

2026-02-20 16:29:06 +02:00

bootstrap_cross_signing.py

feat: Add cross-signing bootstrap + canonicaljson dep

2026-02-15 08:19:38 +02:00

bot.py

fix: increase voice PDF context to 40k chars, fix language detection sanity

2026-02-23 12:40:13 +02:00

docker-compose.yml

fix: Use LITELLM_MASTER_KEY for memory service

2026-02-22 17:25:16 +02:00

Dockerfile

fix(e2ee): revert incorrect HKDF patch, remove pre-ratcheting

2026-02-22 21:44:11 +02:00

Dockerfile.bot

feat(e2ee): Add HKDF E2EE support for Element Call compatibility

2026-02-20 16:29:06 +02:00

e2ee_patch.py

feat(e2ee): Add HKDF E2EE support for Element Call compatibility

2026-02-20 16:29:06 +02:00

patch_sdk.py

feat(e2ee): Add HKDF E2EE support for Element Call compatibility

2026-02-20 16:29:06 +02:00

requirements.txt

feat(voice): all file types + images in voice context (MAT-10)

2026-02-23 11:45:54 +02:00

voice.py

fix: increase voice PDF context to 40k chars, fix language detection sanity

2026-02-23 12:40:13 +02:00

Description

Matrix AI voice agent - LiveKit + LiteLLM

5.5 MiB

Languages

Python 99.1%

Dockerfile 0.8%

Shell 0.1%