Fetching metadata from the HF Docker repository... 
Upload config.py
8837d81 verified - api_integrations fix: lock timeout 30s + training worker + PostgreSQL setup
- skills fix: remove double /v1 from AnyAPI endpoint paths + add missing AOA pairs to currency fallback
- 4.22 kB fix: timer skill, LSTM speakers DB, persona datetime, image gen fallback
- 2.94 kB fix: lock timeout 30s + training worker + PostgreSQL setup
- 5.84 kB fix: 5 broken AnyAPI skills + system prompt 57 skills + twitter import + PDF prompt fix
- 3.65 kB feat: AnyAPI (10 skills) + FastRouter (6+2 rotation) integration
- 679 kB Upload api.py
- 25.3 kB fix: openrouter nemotron→gemini-2.0, mistral timeout 20s+connect, groq payload truncate, factual question gate
- 5.39 kB fix: lock timeout 30s + training worker + PostgreSQL setup
- 14 kB fix: remover Listen Engine gate do /api/akira + remover dedup cache
- 11.6 kB fix: inject mensagem_citada into ThinkingEngine historico_recente para contexto de reply
- 42 kB feat: MAC Drive System + proactive mode + agent_timeout fix
- 5.08 kB feat: remove ALL dedup/anti-loop logic from Python
- 203 Bytes fix: lock timeout 30s + training worker + PostgreSQL setup
- 55.9 kB fix(chat+hf): fluidez conversacional sem busca + HF tier com quota real por conta
- 6.5 kB fix: thread-safety - race conditions, client mutation, global semaphore
- 10.7 kB fix: OpenRouter Nemotron embeddings correct model ID + timezone UTC+1 + web search improvements
- 6.53 kB fix(image): reduce AI artifacts and reorder factory for best quality
- 5.99 kB fix(boot): ciclo infinito em get_current_api_token pendurava o init da API
- 38 kB fix(vision+reply+tokens): HF router free Qwen2-VL-7B, LSTM reply lock, max token largo 8192
- 138 kB Upload config.py
- 28.6 kB Fix: Finalize isolation and continuity boost adjustments
- 19.2 kB feat(session): implementar memória de sessão persistente
- 21.5 kB fix: force float conversion for timestamps from PostgreSQL — prevent TypeError '>' str vs int
- 38.6 kB fix(memory): conversation_id em saves, fallback por numero, pool de conexoes PG
- 59.7 kB fix(memory): conversation_id em saves, fallback por numero, pool de conexoes PG
- 94 kB fix(chat): anti-loop de saida (_collapse_repetition) + busca factual alargada + PG retry apos cooldown
- 24.1 kB fix: debate stance detection + expanded emotional context valid_emotions
- 3.01 kB fix: lock timeout 30s + training worker + PostgreSQL setup
- 9.59 kB feat: MAC Drive System + proactive mode + agent_timeout fix
- 8.14 kB fix: update SoftEdge website URL to softedge-corporation.up.railway.app
- 8.97 kB fix: respostas longas + delírio + fulano vs sicrano (terceira pessoa)
- 3.68 kB perf(ext-gpu): max 180 tokens + deteccao de tunel morto (ngrok 404)
- 3.54 kB fix: FastRouter rotation uses correct env vars (6 provider + 2 CoT keys)
- 3.74 kB fix: FastRouter rotation uses correct env vars (6 provider + 2 CoT keys)
- 49.7 kB fix(chat): chat nunca parte no Gradio (fallback GPU->sem GPU em 429 ZeroGPU) + NLP/classificadores forcados a CPU + probes de CUDA reais
- 8.34 kB fix: openrouter nemotron→gemini-2.0, mistral timeout 20s+connect, groq payload truncate, factual question gate
- 15.8 kB fix(chat): chat nunca parte no Gradio (fallback GPU->sem GPU em 429 ZeroGPU) + NLP/classificadores forcados a CPU + probes de CUDA reais
- 28.2 kB fix: invoice skill parameter mapping - accept company/to aliases
- 3.33 kB fix(hf): snapshot_download autenticado via rotacao de tokens + hf-transfer (fim do 429) + fix Gradio JSON tuple
- 7.81 kB fix(chat+hf): fluidez conversacional sem busca + HF tier com quota real por conta
- 13.8 kB fix: lock timeout 30s + training worker + PostgreSQL setup
- 33.4 kB fix(persona): tom serio-coloquial-logico-racional + fim de respostas com aspas
- 3.95 kB fix: agent_timeout circuit breaker + graceful degradation + MAC v3 + multi-agent orchestrator
- 13.1 kB fix(contexto): dedup de contexto, isolamento por conversa e placeholder de busca
- 33.4 kB fix(jev): continuidade de topico — historico no preclassify + continuity boost System 2 + injecao de encadeamento no prompt
- 12 kB fix(jev): continuidade de topico — historico no preclassify + continuity boost System 2 + injecao de encadeamento no prompt
- 25.1 kB fix(api): NameError explicit_mention_pattern em _build_context_isolation_layer
- 18.2 kB fix: multi-mention context (kiami akira disse X → msg is to kiami), filter observed bleed, ban E contigo
- 21.4 kB fix(generate): BatchEncoding->input_ids/attention_mask explicitos (fim do 500/AttributeError shape)
