duffyduckandClaude Opus 4.8 dc043ceb4d feat(metrics): lokaler LLM-Verbrauch + Claude-Ersparnis in Diagnostic ...

metrics.jsonl-Eintraege tragen jetzt 'source' (claude|local|fast-path).
- log_local_call(): echte usage-Tokens vom Adapter (prompt/completion),
  sonst chars/4-Schaetzung. Geloggt pro Tool-Runde im lokalen Fast-Lane.
- log_fast_path(): reiner Skill, 0 Prompt-Tokens — gesparter Claude-Call.
- aggregate() liefert zusaetzlich by_source (calls/tokens_in/tokens_out).
  Alt-Eintraege ohne source zaehlen als claude (rueckwaerts-kompatibel).

Diagnostic Gehirn-Tab: neue Card "Lokales LLM & Claude-Ersparnis" — pro
Fenster (1h/5h/24h/30d) gesparte Claude-Calls (local + fast-path) und
lokale Token-Last (eigene HW, kein Quota) + Info-Block.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

2026-07-11 16:39:02 +02:00

..

memory

feat(brain): Anhaenge an Memory-Eintraege (Stufe A — Backend)

2026-05-13 02:27:02 +02:00

agent.py

feat(metrics): lokaler LLM-Verbrauch + Claude-Ersparnis in Diagnostic

2026-07-11 16:39:02 +02:00

background.py

feat(tts): System-Flag speak (ja/nein) pro Antwort statt <voice>-Tag-Hack

2026-07-11 14:55:12 +02:00

clean_poisoned_turns.py

fix(brain): Gift-Waechter — Identity-Breaks nie persistieren (Kaskade unterbunden)

2026-07-11 12:25:45 +02:00

conversation.py

feat(projects): Threads im Hauptchat verankert (Stefan-Konzept)

2026-06-13 13:51:26 +02:00

Dockerfile

fix(brain): CPU-only torch — verhindert 5 GB CUDA-Bloat im Brain-Image

2026-05-23 15:45:51 +02:00

local_llm.py

feat(local-llm): B0.5 — llama-swap + lokale Modellauswahl in Diagnostic

2026-07-11 14:11:59 +02:00

main.py

feat(tts): System-Flag speak (ja/nein) pro Antwort statt <voice>-Tag-Hack

2026-07-11 14:55:12 +02:00

memory_attachments.py

feat(brain): memory_save mit attach_paths — ARIA haengt Bilder selbst an

2026-05-13 02:57:02 +02:00

metrics.py

feat(metrics): lokaler LLM-Verbrauch + Claude-Ersparnis in Diagnostic

2026-07-11 16:39:02 +02:00

migrate_backfill_projectid.py

fix(projects): Backfill v2 — robuster Match trotz GPS/Barge-In/FILE-Marker

2026-07-03 02:34:17 +02:00

migration.py

feat(brain): Phase B — Vector-DB-Memory, Conversation-Loop, Skills, Tool-Use

2026-05-11 22:23:17 +02:00

oauth.py

feat(oauth): ARIA kann Provider selbst registrieren + Custom-Provider in Diagnostic & App

2026-05-24 20:16:31 +02:00

projects.py

feat(projects): Threads im Hauptchat verankert (Stefan-Konzept)

2026-06-13 13:51:26 +02:00

prompts.py

fix(privacy): harte Diskretions-Regel — intime Details NIE ungefragt preisgeben

2026-07-11 15:38:52 +02:00

proxy_client.py

feat(multitask): kontext-getaggte Activity + kontext-scoped Cancel (Restpunkte 1+2)

2026-07-10 21:29:59 +02:00

requirements.txt

feat(brain): Phase B — Vector-DB-Memory, Conversation-Loop, Skills, Tool-Use

2026-05-11 22:23:17 +02:00

router.py

fix(router): lokales Tier eskaliert Gedaechtnis-/Personen-Fragen an Claude

2026-07-11 15:51:18 +02:00

seed_rules.py

refactor(brain): Fast-Path als Skill-Capability — fast_patterns im Manifest

2026-06-06 08:56:15 +02:00

skill_templates.py

feat(brain): skill_scaffold — Templates statt Skill aus dem Nichts

2026-05-30 00:02:45 +02:00

skills.py

refactor(brain): Fast-Path als Skill-Capability — fast_patterns im Manifest

2026-06-06 08:56:15 +02:00

triggers.py

feat(brain+ui+app): Triggers — passive Aufweck-Quellen fuer ARIA

2026-05-12 00:38:58 +02:00

watcher.py

feat(trigger): entered_near + left_near — drei Modi fuer near()-Watcher

2026-05-14 18:29:39 +02:00