# Token-Savior Memory Export — 222 memorias relevantes # Generado: 2026-04-27 — workspace F:/gaston-workspace ====================================================================== TITLE: Auto-guardado obligatorio de memoria TYPE: guardrail | IMPORTANCE: 10 TAGS: ['meta', 'memoria', 'automatizacion'] Al cerrar cualquier unidad de trabajo (bugfix, decision, convencion, API key, URL, comando util, descubrimiento), guardar memoria ANTES de resumir al usuario. No preguntar permiso: guardar y mencionar que quedo registrado. Why: Gaston no quiere que me olvide de guardar nada nunca. Con auto_capture del mode code ampliado + esta regla, cubro todos los canales (tool calls + conclusiones en lenguaje natural). How to apply: Antes de terminar una respuesta que contenga info reutilizable: memory_db.observation_save(...) con type apropiado, title corto, why + how_to_apply. Importance >=7 si es preferencia personal, >=9 si es guardrail permanente. ====================================================================== TITLE: CannTrace C7 detalles finales (fechas, cuadros, bolsas) + C8 no existe TYPE: context | IMPORTANCE: 10 TAGS: ['canntrace', 'c7', 'trazabilidad', 'fixes', 'prompt-nueva-sesion'] Cierre sesion 25/04/2026 noche. Fixes finales en C7: 1) Cosecha 192 plantas (no 63 que era de C9). Filtra cosechaReal por camada estricta. 2) Secado 14 dias reales (era 574 por fallback 2025-01-01). Helper fechasValidas() filtra el sentinel. Usa datos_extra.fecha_fin_secado de CM-RE-0205. 3) Secado detail: "8 cuadros (CDS8/9/10/11/19/20/21/22) · SFL2 · 192 plantas". 4) Fraccionamiento: "75 bolsas (~400g) · 30.179g totales · 5 lotes paqueteria". 5) Deposito Final: "75 bolsas · 30.179g · ~400g por bolsa · 138 codigos comerciales · DEPOSITO FIS". Codigo: 25.FIS.BI1.C7.8 -> 25.FIS.BI5.C7.22. C8 confirmado NO existe como camada productiva: solo 1 registro en diccionario CM-RE-0210. Camadas reales: C7, C9, C11, C12, C15, C16. Prompt nueva sesion guardado en F:\gaston-workspace\memory‚6-04-25-canntrace-prompt-nueva-sesion.md Why: Datos de C7 tenian inconsistencias visuales (63 plantas de C9, 574 dias de fallback, etc). Ahora cadena coherente con datos reales del Excel maestro. How to apply: Ver C7: https://canntrace.pages.dev/trazabilidad → click C7 → click cualquier stage. Para nueva sesion usar el prompt en 2026-04-25-canntrace-prompt-nueva-sesion.md ====================================================================== TITLE: LightRAG migrado a embeddings local (TEI sentence-transformers) TYPE: decision | IMPORTANCE: 10 TAGS: ['lightrag', 'embeddings', 'tei'] Antes: Gemini embed-001 (1536 dim, 1000/dia free, hit cuota). Ahora: TEI HuggingFace text-embeddings-inference cpu-1.7 + sentence-transformers/all-MiniLM-L6-v2 (384 dim, sin cuotas, ~50ms/embed local CPU). Container embedding-server:8090 con --auto-truncate --max-batch-tokens 16384. .env LightRAG: EMBEDDING_BINDING=openai EMBEDDING_BINDING_HOST=http://embedding-server:80 EMBEDDING_DIM=384 EMBEDDING_TOKEN_LIMIT=512 EMBEDDING_BATCH_NUM=8 EMBEDDING_FUNC_MAX_ASYNC=4. Backups del store viejo en rag_storage/gaston.OLD.* y gaston.PREMIGRACION.* y gaston.PRE-LOCAL2.* Why: Sin cuotas, mas rapido, privado, no rompe por rate limits How to apply: Si Gemini cuota issue vuelve, ya esta cubierto. Para cambiar modelo: edit compose --model-id + reset embed dim en LightRAG .env + clear vector store (vdb_*.json) ====================================================================== TITLE: Cloudflare setup completo (n8n-nocodedb.info) 2026-04-27 TYPE: context | IMPORTANCE: 10 TAGS: ['cloudflare', 'tunnel', 'dns'] Account 5e06901c2116743475d5c5d379044995. Domain n8n-nocodedb.info zone ce8c52976a18f89539978a1973f5d084 plan Free, expira 2027-05-09 (renueva via Hostinger). Tunnel office-5090 id eb337efc-2da2-45af-8be2-4c54164b6eb7 STATUS INACTIVE (cloudflared no corriendo aun en oficina). Token guardado en F:/gaston-workspace/ai-stack/cloudflared-office-token.txt. 9 subdomains preconfigured ingress: ollama vllm lightrag openwebui hermes mcpo embed tei office-rdp todos -> http://localhost: en oficina. 9 DNS records CNAME proxied a tunnel. 9 Apps Zero Trust Access creadas con policy email-only para gastonrpersoglia@gmail.com. Zone hardened: Always HTTPS ON, Min TLS 1.2, Auto HTTPS Rewrites ON, WebSockets ON. R2 NO activado (necesitaria tarjeta). API token cfut_urABJ9fDgnRtClTks... en config con scopes admin completos. Why: Para activar todo: instalar cloudflared en pc-oficina con el token guardado, los DNS y ACL ya estan listos. How to apply: En PC oficina: winget install cloudflare.cloudflared && cloudflared service install . Acceso publico via https://.n8n-nocodedb.info con auth email. ====================================================================== TITLE: OpenWebUI: WEBUI_SECRET_KEY + builtin_tools fix 2026-04-27 TYPE: bugfix | IMPORTANCE: 10 TAGS: ['openwebui', 'bugfix', 'jwt'] BUG 1: WEBUI_SECRET_KEY se guarda en /app/backend/.webui_secret_key (FUERA del volume) -> cada docker compose --force-recreate genera nuevo secret -> JWTs viejas invalidadas. FIX: persistido en docker-compose.yml as env var WEBUI_SECRET_KEY=ZBqBls7vfn9fAyR/. BUG 2: 9 de 10 modelos favoritos tenian params.function_calling=native que OVERRIDE el global default -> tools nativos query_knowledge_bases se inyectaban -> Groq+Llama returns count como string -> schema mismatch error. FIX: removido params.function_calling de los 10 favoritos + agregado meta.capabilities.builtin_tools=False (doble seguridad). Para generar nuevo JWT: docker exec openwebui python3 -c "import sqlite3,jwt,time; c=sqlite3.connect("/app/backend/data/webui.db"); ...". Why: Estos bugs se vuelven a manifestar facil. El secret_key fix es CRITICO o pierdes tu sesion cada recreate. How to apply: Si despues de un recreate la UI te pide login y tu cuenta no funciona, regenera JWT desde DB con el secret en compose. ====================================================================== TITLE: Stack roadmap: PC oficina + 5090 (proximo sprint) TYPE: decision | IMPORTANCE: 10 TAGS: ['roadmap', '5090', 'tailscale', 'cloudflare'] Pendings: 1) Instalar Tailscale en PC oficina (winget) + login mismo cuenta. 2) Instalar cloudflared en oficina con token de F:/gaston-workspace/ai-stack/cloudflared-office-token.txt. 3) Levantar Ollama o vLLM con Qwen 32B en RTX 5090. 4) Cuando funcione: switchear hermes a HERMES_LLM_PROVIDER=openai HERMES_OPENAI_BASE_URL=http://100.X.Y.Z:8000/v1 (lineas comentadas en docker-compose.yml ya listas). 5) En OpenWebUI: agregar provider apuntando a la IP Tailscale oficina. RDP desde casa via Tailscale (Windows Pro o RustDesk). PC casa NO sirve para LLM (Ryzen 1500X + GPU 2011), solo embeddings ya OK. Why: Roadmap corto cuando vaya a oficina, todo el setup lado casa esta listo. How to apply: Cuando llegue oficina: tailscale up con auth-key + cloudflared service install + Ollama install. Verifica desde casa. ====================================================================== TITLE: polymarket-weather:fee-changes TYPE: guardrail | IMPORTANCE: 10 TAGS: ['polymarket', 'fees', 'stablecoin', 'guardrail', 'migration'] CRITICO 30-marzo-2026: Polymarket agrego fees a weather markets (antes eran free como politics/geopolitics). Ahora hay fee dinamico basado en probabilidad — outcomes seguros pagan menos. INVALIDA simulaciones que asumian fees=0. CRITICO 28-abril-2026: stablecoin upgrade USDC.e → Polymarket USD (PolyUSD), ~1h downtime, address de stablecoin cambia. Why: Calcular EV sin fees actuales = trades perdedores reales. Y aprovals on-chain anteriores quedan invalidas tras upgrade de stablecoin. How to apply: Antes de cada trade real → llamar pm_get_fee_rate(token_id) para obtener fee actual y ajustar EV. Tras 28-abril → detectar nueva address de stablecoin y re-aprobar contratos automaticamente con pmx_approve_post_stablecoin_upgrade. Pausar trading si Polymarket cambia fees nuevamente sin aviso. ====================================================================== TITLE: instructivo-oficina-master-doc TYPE: reference | IMPORTANCE: 10 TAGS: ['instructivo', 'setup', 'oficina', 'tailscale', 'master-doc', 'reference'] Instructivo MAESTRO completo del stack Gaston en F:\gaston-workspace\INSTRUCTIVO_OFICINA.md (.md, .pdf, y MEMORIAS_TOKEN_SAVIOR_EXPORT.txt). Incluye: snapshot del stack actual (containers, MCPs Linux 8770, MCPs Windows 8771, polymarket-weather-mcp), Camino A (Tailscale serve para conectar oficina a casa), Camino B (replicar todo en oficina desde cero), comandos diarios, troubleshooting conocido, paths importantes, proximos pasos. Tailnet: tail89f35a.ts.net, PC casa: pc-casa.tail89f35a.ts.net (100.91.248.66). MCPO Bearer: mcpo-gaston-secret. PDF: 35KB, MD: 25KB. Why: Para que Gaston pueda configurar la PC oficina y conectarla con la PC casa, y para no perder contexto entre sesiones. Documento maestro de referencia. How to apply: Cuando Gaston pregunte por setup, replicacion, conexion casa-oficina, o pierda contexto sobre el stack: leer F:\gaston-workspace\INSTRUCTIVO_OFICINA.md (es el snapshot mas reciente y completo). Si necesita texto sin formato (celu): MEMORIAS_TOKEN_SAVIOR_EXPORT.txt. Si quiere imprimir: INSTRUCTIVO_OFICINA.pdf. ====================================================================== TITLE: Idioma y estilo de Gaston TYPE: convention | IMPORTANCE: 9 TAGS: ['estilo', 'idioma', 'comunicacion'] Responder en espanol rioplatense, terminos tecnicos en ingles. Conciso, directo, con ejemplos cuando aporten, cero relleno. Why: Pedido explicito del usuario; prioriza economia de tokens. How to apply: Evitar resumenes largos; no repetir lo que el diff muestra; verificar codigo antes de declarar listo. ====================================================================== TITLE: CannTrace ingest xlsx CM-RE-1010 v2 completo (24/04/2026 madrugada) TYPE: context | IMPORTANCE: 9 TAGS: ['canntrace', 'ingest', 'xlsx', 'supabase', 'cumcs', 'trazabilidad'] Sesion completa: ingestamos 13.404 raw + 9.961 estructuradas (~26K filas nuevas) del xlsx CM-RE-1010 Matriz Consolidada v2.xlsx en Supabase canntrace. Migrations nuevas: - 20260424_fix_tabla_destino_cumcs (catalogo G01/G02/G03/G06) - 20260424_add_fingerprint_indexes (11 unique indexes parciales + datos_extra a resultados_laboratorio) - 20260424_create_xlsx_raw_rows (tabla captura cruda) Script en scripts/ingest-xlsx-cumcs/ idempotente con fingerprint sha256 + ON CONFLICT DO NOTHING. 28 hojas mapeadas en mappings.yaml. Re-ejecutar con SUPABASE_PAT=sbp_... node ingest.mjs --apply. Trazabilidad enriquecida: PaginaTrazabilidad.tsx con queries paralelas a registros_cosecha + trazabilidad + ambientales + calidad. Stages 1/3/4/5/6/8 muestran datos reales (verificado en C9: 63 plantas PETE HOPE 2025-07-03). Consultas drill-down: PaginaConsultas.tsx con drawer lateral 5xl, paginacion 50/pagina, busqueda libre. Mapa armonico: PaginaMapa.tsx con SVG 9 areas en familia verde+dorado+gris (no arcoiris), modal lateral con CUMCS asociados. Why: Antes el xlsx estaba en Storage pero la app no lo leia. Trazabilidad solo usaba lotes (sin JOIN). Catalogo CUMCS apuntaba mal a tablas genericas para 4 grupos. How to apply: Re-ingestar: cd scripts/ingest-xlsx-cumcs && SUPABASE_PAT=sbp_b8403ab25dd058e9334143e32bea1fca5c7bde75 node ingest.mjs --apply. Para nuevas hojas: agregar entry en mappings.yaml con header_keywords + columns map. ====================================================================== TITLE: ai-stack docker-compose optimizado 2026-04-27 TYPE: context | IMPORTANCE: 9 TAGS: ['ai-stack', 'docker', 'config'] F:/gaston-workspace/ai-stack/docker-compose.yml — 5 services healthy: openwebui:0.9.2 (3000), lightrag:v1.4.15 (9622), mcpo:main (8770), hermes:gaston/hermes-shim:latest (8642), embedding-server:cpu-1.7 (8090). Todos con healthcheck cada 30s + log rotation 10MB x 3 archivos. WEBUI_SECRET_KEY=ZBqBls7vfn9fAyR/ persistido en compose para que JWTs sobrevivan recreates. Volumes: openwebui_data, hermes_data (external), F:/gaston-workspace/gaston-trading/tools/LightRAG/rag_storage, F:/gaston-workspace/ai-stack/embedding-models. Why: Recordar el estado optimo del stack para no romper en futuras sesiones. How to apply: Cualquier cambio al stack: cd F:/gaston-workspace/ai-stack && docker compose up -d --force-recreate ====================================================================== TITLE: 6 admin MCPs custom + skills (2026-04-27) TYPE: context | IMPORTANCE: 9 TAGS: ['mcp', 'admin', 'stack'] Stack admin custom (Python FastMCP en C:/tools/-mcp/server.py): hermes-admin (72 tools), mcpo-admin (36 multi-instance container+host), openwebui-admin (67), tailscale-admin (66), cloudflare-admin (87), hostinger-admin (72). Skills en C:/Users/Gaston/.claude/skills//SKILL.md. Configurados en C:/Users/Gaston/AppData/Roaming/Claude/claude_desktop_config.json con env vars de creds. Why: Cuando necesite admin de cualquier capa del stack, usar el MCP correspondiente en vez de curl/dashboard How to apply: En Claude Code llamar mcp__-admin__. health() en cada uno para verificar. ====================================================================== TITLE: Tailscale tailnet state 2026-04-27 TYPE: context | IMPORTANCE: 9 TAGS: ['tailscale', 'vpn', 'setup'] Cuenta gastonrpersoglia@gmail.com. Tailnet tail89f35a.ts.net. UN solo device: pc-casa (id 1556119855529414, IP 100.91.248.66, tag:home). MagicDNS ON. ACL con tagOwners definidos para tag:home y tag:server (vacio aun, se usa cuando llegue PC oficina). Settings: device_key_duration 90d, HTTPS ON, auto-updates ON. accept-routes=true en pc-casa. PC casa NO soporta SSH server (Windows). Para futuro PC oficina: tskey-auth-... existe (reusable, en config) o crear single-use con preauthorized=true tags=tag:server. Why: Toda futura admin del tailnet desde aca via mcp__tailscale-admin__*. Cuando llegue PC oficina: tag:server + restringir ACL. How to apply: Para registrar pc-oficina sin browser: tailscale up --authkey= --advertise-tags=tag:server --hostname=pc-oficina ====================================================================== TITLE: LightRAG workspace gaston: 22 skills indexadas + bridge filter TYPE: context | IMPORTANCE: 9 TAGS: ['lightrag', 'rag', 'skills'] Workspace gaston en LightRAG (NetworkX + NanoVectorDB local). 22 docs indexados: 14 admin skills (desktop-commander, memory, excel, supabase, token-savior-recall, lightrag, openwebui, openwebui-admin, MCP_DOCKER, ruflo, context7, hermes, hermes-admin, mcpo-admin) + bingx-trading + 4 pine (pirenest, sz-custom, wedges-sr, terminator) + 3 admin nuevos (tailscale-admin, cloudflare-admin, hostinger-admin). NO incluye: tradingview*, telegram*, ccxt-mcp (trading skills, no subidas todavia). Filter en OpenWebUI lightrag_kb_bridge intercepta cuando se attacha la KB Gaston KB (LightRAG mirror, id 564f45f4-9760-49fa-ab86-e0a6f2bffecb), strip los archivos del body, query LightRAG /query mode=hybrid top_k=40 only_need_context=true, inyecta resultado como system message. Why: El bridge es CRITICO: el dropdown KB de OW dice 0 archivos pero el filter consulta LightRAG real con todo lo indexado. How to apply: En chat OW: + -> Conocimiento -> Gaston KB (LightRAG mirror) -> mandar query. Filter dispara automatico (visible en logs OpenWebUI con [LightRAG_Bridge]). ====================================================================== TITLE: polymarket-weather:risk-config-aggressive TYPE: convention | IMPORTANCE: 9 TAGS: ['polymarket', 'weather', 'aggressive', 'bankroll', 'kelly', 'safety'] Modo Aggressive elegido por Gaston con bankroll $100. Half Kelly (0.5 fraction). MAX_PCT_PER_TRADE=0.15 (15% bankroll vivo). HARDCAP_PER_TRADE_USD=20 absoluto. DAILY_STOP_LOSS_PCT=0.25 (-25% dia → pausa 24h). DRAWDOWN_PAUSE_PCT=0.40 (-40% acumulado → pausa 2 semanas). MIN_EV=0.05. DRIFT_PAUSE_THRESHOLD=0.50 (rolling 30 trades, win rate <50% → pausa auto). AUTO_SKIP_NO_GAP_DAYS=true. ENABLE_LADDERING=true. ENABLE_NEGRISK=true. ENABLE_HODL=false. SCAN_CATEGORIES=weather,economy,oil,negrisk. TRADING_MODE=paper minimo 7 dias antes de live. Why: Configuracion calibrada matematicamente para $100 bankroll. Full Kelly arruina con bankroll chico. Half Kelly = 75% growth con 50% varianza. Triple safety hardcoded para evitar ruina. How to apply: Aplicar al .env del MCP polymarket-weather. NUNCA desactivar las reglas de safety hardcoded. NUNCA invocar pmx_place_order_* sin confirm=True. ====================================================================== TITLE: polymarket-weather:isp-dns-hijacking TYPE: warning | IMPORTANCE: 9 TAGS: ['polymarket', 'dns', 'isp', 'hijacking', 'blocker', 'fibertel'] Fibertel (ISP de Gaston en AR) intercepta DNS para *.polymarket.com agregando sufijo falso .com.ar (resuelve a 208.115.249.235 en lugar del CDN real). Afecta gamma-api.polymarket.com, data-api.polymarket.com, clob.polymarket.com. NOAA (aviationweather.gov) y Open-Meteo NO afectados. Why: El MCP polymarket-weather no puede resolver dominios de Polymarket. No es bug del codigo — es DNS hijacking del ISP. How to apply: Soluciones: (1) Cambiar DNS de Windows a Cloudflare 1.1.1.1 + 1.0.0.1 o Google 8.8.8.8 (Configuracion → Red → Adaptador → IPv4 → DNS manual). (2) Usar Tailscale exit node fuera de Argentina. (3) Agregar entries a C:\Windows\System32\drivers\etc\hosts con IPs de Cloudflare CDN. (4) Implementar DNS-over-HTTPS en util/http.py como fallback. Recomendado opcion (1) — afecta toda la PC pero es el unico fix permanente. ====================================================================== TITLE: polymarket-mcp:architecture-sse-via-mcpo TYPE: convention | IMPORTANCE: 9 TAGS: ['mcp', 'mcpo', 'sse', 'docker', 'architecture', 'fastmcp', 'transport'] polymarket-weather-mcp NO se registra en mcpo-config.json como type=stdio porque MCPO corre en container Linux y no puede ejecutar C:\Python314\python.exe. Patron correcto: el MCP corre en host Windows como SSE server (FastMCP transport=sse, bind 0.0.0.0:8779), MCPO se conecta via type=sse y url=http://host.docker.internal:8779/sse. Importante: FastMCP por default tiene TransportSecuritySettings con allowed_hosts=[127.0.0.1, localhost] que rechaza host.docker.internal con HTTP 421 — solucion: pasar TransportSecuritySettings(enable_dns_rebinding_protection=False, allowed_hosts=[*], allowed_origins=[*]) al constructor FastMCP. Why: Si en algun futuro alguien intenta registrar otro MCP custom Python en MCPO via stdio con C:\Python314, va a fallar igual. Esta es la convencion correcta en este setup. How to apply: Para nuevos MCPs custom Python: (1) En el server.py agregar TransportSecuritySettings con allowed_hosts=[*]. (2) Soportar argparse --transport sse --host 0.0.0.0 --port N. (3) En mcpo-config.json usar {type: sse, url: http://host.docker.internal:N/sse}. (4) Lanzar el server.py separado como proceso Windows (scheduled task ONSTART o manual). (5) El API key MCPO actual: mcpo-gaston-secret (Authorization: Bearer mcpo-gaston-secret). ====================================================================== TITLE: mcpo-architecture-dual-instance TYPE: convention | IMPORTANCE: 9 TAGS: ['mcpo', 'docker', 'windows', 'architecture', 'dual-instance'] El stack MCPO tiene DOS instances: (1) Container Linux Docker en puerto 8770 (config: F:\gaston-workspacei-stack\mcpo-config.json) para MCPs npx/uvx que corren dentro del container. (2) Host Windows nativo en puerto 8771 (config: F:\gaston-workspacei-stack\mcpo-windows-config.json, lanzado por start-mcpo-windows.ps1) para MCPs Python con dependencias Windows. Auth bearer key compartido: mcpo-gaston-secret. OpenWebUI los consume via http://mcpo:8000/ para Linux y http://host.docker.internal:8771/ para Windows. Why: Si registro un MCP custom Python Windows (path C:\Python314\python.exe) en el config Linux (8770), MCPO container no puede ejecutarlo. Hay que usar el config Windows (8771). How to apply: Para nuevos MCPs Python Windows: agregar entry stdio en mcpo-windows-config.json con command absoluto al python.exe + args con path absoluto al server.py. Reiniciar MCPO Windows con: powershell -ExecutionPolicy Bypass -File F:\gaston-workspacei-stack\start-mcpo-windows.ps1 (kill primero el proceso si esta corriendo). Para MCPs Linux/Docker (npx/uvx) usar mcpo-config.json y docker restart mcpo. ====================================================================== TITLE: Proyectos activos del workspace TYPE: context | IMPORTANCE: 8 TAGS: ['proyectos'] Proyectos en F:/gaston-workspace: cannabis-trazabilidad (CannTrace), gaston-trading + codex-trading-plugin, consultora-ojeda-araujo (GAMP5), plataforma-cursos. How to apply: Ante pedido ambiguo, identificar a que proyecto pertenece antes de actuar. ====================================================================== TITLE: Memoria persistente via token-savior TYPE: convention | IMPORTANCE: 8 TAGS: ['memoria', 'tokens', 'infraestructura'] Sistema token-savior instalado con hooks completos (SessionStart inyecta indice adaptativo, SessionEnd/Stop cierran sesion, PreCompact preserva contexto). DB SQLite en ~/.local/share/token-savior. MCP tools mcp__token-savior-recall__* para consultar/guardar. How to apply: Para persistir algo entre sesiones: usar memory_save en vez de reescribir archivos manualmente. SessionStart lo traera de vuelta automaticamente. ====================================================================== TITLE: Hermes shim ES Hermes Agent real (no proxy) TYPE: context | IMPORTANCE: 8 TAGS: ['hermes', 'docker', 'api'] gaston/hermes-shim:latest extiende nousresearch/hermes-agent:latest y agrega FastAPI (shim.py). El binario hermes esta en /opt/hermes/.venv/bin/hermes. Cuando POST /v1/chat/completions llega al shim, ejecuta /opt/hermes/.venv/bin/hermes chat -q "" --yolo --max-turns 3 --ignore-rules y captura stdout. Skills bundled (73 categorias) las maneja el binario hermes mismo. Memory en /opt/data/memory/. Latencia ~80s por request porque hace pipeline completo. NO es un proxy a OpenRouter. Why: Aclarar la confusion. Antes pense que era proxy tonto, NO. Tiene hooks, skills, memoria, personality real. Solo un wrapper HTTP encima. How to apply: Para usar Hermes via API: POST http://localhost:8642/v1/chat/completions con model=hermes-agent. O via mcp__hermes-admin__chat. Modificar provider/model: hermes-admin set_provider/set_model. ====================================================================== TITLE: polymarket-weather:wallets-verified TYPE: context | IMPORTANCE: 8 TAGS: ['polymarket', 'weather', 'wallets', 'on-chain'] Top traders verificados ON-CHAIN en Polymarket (los unicos confiables como benchmark). gopfan2: 0xf2f6af4f27ec2dcf4072095ab804016e14cd5817 (+1-2M USD, weather scan buy<0.15 sell>0.45). Sharky6999: 0x751a2b86... (+480K-934K, HODL high-prob entrada 99.67c, 12k+ tx, 97-98% win rate). Theo4: 0x56687bf... (+22.05M, politica sistemica, 88.9% win rate). Hardcodear estas 3 en core/chain/known_traders.py del MCP. Why: Solo estas 3 wallets son verificables en Polygonscan/polymarketanalytics. ColdMath/1pixel/meropi son blogs sin verificacion on-chain. How to apply: Cuando el MCP polymarket-weather necesite track_top_traders o como benchmark de estrategias, usar SOLO estas 3 addresses. Para track de cambios de estrategia, alertas trimestrales. ====================================================================== TITLE: windows-dns-original-backup TYPE: context | IMPORTANCE: 8 TAGS: ['dns', 'windows', 'backup', 'fibertel', 'cloudflare', 'rollback'] Backup del DNS original de Windows ANTES del cambio a Cloudflare (2026-04-27). Interface Ethernet IPv4 DNS: 181.30.140.195, 181.30.140.135 (Fibertel — DNS que hijackea polymarket.com). Tailscale: fec0:0:0:ffff::1, fec0:0:0:ffff::2, fec0:0:0:ffff::3 (MagicDNS, NO tocado). Otros adapters (vEthernet WSL, Default Switch) tampoco tocados. Why: Si el cambio a Cloudflare 1.1.1.1 rompe algo (acceso a servicios LAN nombrados, captive portals, etc), poder revertir. How to apply: Para rollback ejecutar como Admin en PowerShell: Set-DnsClientServerAddress -InterfaceAlias "Ethernet" -ResetServerAddresses; ipconfig /flushdns. O correr F:\gaston-workspace\polymarket-weather-mcp\scripts ollback_dns.ps1. ====================================================================== TITLE: openwebui-tool-server-add-via-db TYPE: convention | IMPORTANCE: 8 TAGS: ['openwebui', 'tool-server', 'db', 'sqlite', 'config'] Para agregar un tool server a OpenWebUI sin la UI: editar webui.db tabla config (id=1, campo data JSON) → tool_server.connections (lista). Cada entry: {url, path:openapi.json, type:openapi, auth_type:bearer, key:mcpo-gaston-secret, info:{name,description}, config:{enable:true}}. Despues docker restart openwebui para que recargue. Path al sqlite dentro del container: /app/backend/data/webui.db. Admin user existe (gastonrpersoglia@gmail.com, role=admin) pero la API key no esta en la tabla user — esta en la tabla api_key. Why: El MCP openwebui-admin a veces se desconecta. Tener el patron via DB es backup directo. How to apply: Script de referencia: F:\gaston-workspace\polymarket-weather-mcp\scriptsdd_owui_tool_server.py. Hacer docker cp del script al container y docker exec python. ====================================================================== TITLE: openwebui-function-calling-native TYPE: convention | IMPORTANCE: 8 TAGS: ['openwebui', 'function-calling', 'tools', 'config', 'json'] OpenWebUI por default tiene models.default_params.function_calling=default — esto hace que cuando un tool MCPO devuelve JSON, el LLM lo MUESTRE crudo en lugar de procesarlo. Cambiar a native (en webui.db config tabla, campo data JSON) para que el tool output se inyecte como mensaje de tool en el flujo y el LLM razone arriba en lenguaje natural. Why: User Gaston experimentaba que TODOS los modelos (Llama 3.3 70B, etc.) devolvian JSON crudo cuando invocaban tools del MCP polymarket-weather. Era setting global de OpenWebUI, no del MCP ni del modelo. How to apply: Para arreglar: docker exec openwebui python con sqlite3 → SELECT data FROM config WHERE id=1 → modify json.models.default_params.function_calling = native → UPDATE → docker restart openwebui. Script de referencia: F:\gaston-workspace\polymarket-weather-mcp\scripts ix_function_calling.py ====================================================================== TITLE: Identidad: Sophie TYPE: convention | IMPORTANCE: 7 TAGS: ['identidad'] Gaston llama al asistente "Sophie" (bautismo 2026-04-22). Responder al nombre cuando lo use, sin forzarlo. ====================================================================== TITLE: Setup Obsidian + junctions de memoria TYPE: context | IMPORTANCE: 7 TAGS: ['obsidian', 'memoria', 'setup'] Dos vaults Obsidian: (1) F:/gaston-workspace/llm-wiki/Gastoon Workspace, (2) F:/gaston-workspace/trading-wiki. Junctions: C:/Users/Gaston/.claude/projects/F--gaston-workspace/memory y /Claude-Memory -> F:/gaston-workspace/memory (fuente unica). How to apply: Escribir memorias en F:/gaston-workspace/memory/*.md aparece automaticamente en ambos vaults como Claude-Memory. ====================================================================== TITLE: Hostinger account state 2026-04-27 TYPE: context | IMPORTANCE: 7 TAGS: ['hostinger', 'vps', 'dns'] gastonrpersoglia@gmail.com. 1 dominio (n8n-nocodedb.info, expira 2027-05-09). 0 VPS contratado. 0 cards on file. Token API guardado en config MCP hostinger-admin: idaaReFhiOypX2lzEx7kJldejxg4ek9l1Wr44FeBceb40589. Scopes: billing, domains, DNS, VPS lifecycle, snapshots, backups OK. Sin scope post-install ni firewall (esos requieren tier mayor o flag adicional). Datacenters disponibles: bos2 (Boston, mas cerca de Argentina ~120ms). Templates VPS 117 (Ubuntu/Debian/AlmaLinux/Rocky + with-app). Why: Si necesita VPS para correr cosas 24/7 (ej Ollama si no tiene 5090), el setup de hostinger esta listo via MCP. How to apply: mcp__hostinger-admin__* para todo. health() para verificar. ====================================================================== TITLE: MCP_DOCKER reducido a 1 server por timeout en Claude Code TYPE: bugfix | IMPORTANCE: 7 TAGS: ['mcp_docker', 'docker', 'timeout'] Antes: docker mcp gateway run con 4 servers (filesystem, github, puppeteer, sequentialthinking) tardaba 3-15s en init -> Claude Code timeout y desconectaba. Fix: docker mcp server disable filesystem/github/puppeteer (mantengo solo sequentialthinking). Ahora ~6s init, 7 tools (1 + 6 internos del gateway). Para re-habilitar: docker mcp server enable . Why: Cuando alguno de esos MCPs falte, no es bug del MCP — esta deshabilitado a proposito. How to apply: docker mcp server list. Si necesito puppeteer, prefiero claude-in-chrome. Si necesito filesystem, prefiero desktop-commander. GitHub: enable solo cuando lo use. ====================================================================== TITLE: polymarket-weather:wallets-not-verified TYPE: warning | IMPORTANCE: 7 TAGS: ['polymarket', 'wallets', 'verification', 'fact-check'] ColdMath (~104K USD), 1pixel (18.5K de 2.3K depositados), meropi (~30K micro-bets) son menciones en blogs/Medium. Sus wallet addresses NO se encontraron publicamente. NO usar como benchmark cuantitativo. Why: El simulador anterior (otra IA) y articulos de Medium asumen como verdad cifras de estos traders. Si las repetimos sin verificar somos credulos. Solo gopfan2/Sharky6999/Theo4 estan confirmados. How to apply: Si Gaston pregunta por ColdMath/1pixel/meropi, aclarar que son no verificables. NO incluir en core/chain/known_traders.py del MCP. Si aparecen sus addresses publicas en el futuro, recien ahi agregar. ====================================================================== TITLE: polymarket-weather:edge-erosion-status TYPE: context | IMPORTANCE: 7 TAGS: ['polymarket', 'edge', 'market-state', 'strategy'] Estado del edge en Polymarket abril 2026: WEATHER spreads se comprimieron de 3-5c (2024) a <1c (2026), edge erosionado por bots. ECONOMY (162 mercados, CPI/Jobs/Fed) tiene edge medio. OIL/COMMODITIES (153 mercados) sub-explotada post-crisis Hormuz Q1 2026 — pocos bots aun. NEGRISK rebalancing genero 29M de 39.59M arb total = 73% del profit en 8.6% de oportunidades = 29x ventaja eficiencia. Arbitraje logico (sum>1) duraba 30-60s en 2024, ahora 2.7s — manual imposible. 84% de traders pierden dinero, solo 0.033% top captura mayoria de profits. Why: Saber donde esta el edge real ahora vs donde estaba antes evita estrategias obsoletas. Weather puro ya no rinde como en 2024-2025; hay que combinarlo con economy + oil + negrisk. How to apply: El MCP polymarket-weather scan_categories debe ser weather+economy+oil+negrisk (no solo weather). Priorizar oil/commodities y negrisk para mayor edge. Re-evaluar trimestralmente. ====================================================================== TITLE: windows-dns-fix-applied TYPE: context | IMPORTANCE: 7 TAGS: ['dns', 'cloudflare', 'windows', 'fix-applied'] DNS de Windows cambiado a Cloudflare 1.1.1.1 + 1.0.0.1 + Google 8.8.8.8 el 2026-04-27 (interface Ethernet). Confirmado: gamma-api.polymarket.com resuelve correctamente a Cloudflare CDN (172.64.153.51, 104.18.34.205). Polymarket APIs accesibles. Tailscale, WSL, LightRAG, Hermes Docker NO afectados. Why: Fibertel hijackea *.polymarket.com con sufijo .com.ar — el cambio a DNS publicos lo evita. How to apply: Si en algun momento necesitas revertir, ejecutar como Admin: Set-DnsClientServerAddress -InterfaceAlias "Ethernet" -ResetServerAddresses; ipconfig /flushdns. O correr scripts/rollback_dns.ps1. ====================================================================== TITLE: Estructura vault Gastoon Workspace TYPE: context | IMPORTANCE: 6 TAGS: ['obsidian', 'vault', 'estructura'] Vault LLM llenado con estructura: index.md (home), Daily/, Inbox/, Proyectos/ (CannTrace, Trading, Consultora, Cursos), Referencias/, Templates/ (daily, meeting, project), Archivo/, 00-Index/. Claude-Memory y Claude-Memory-DB son junctions. How to apply: Al escribir notas en este vault usar frontmatter YAML con tags. Los hubs de Proyectos/ ya linkean a los Proyecto-*.md de Claude-Memory. ====================================================================== TITLE: Proyecto: Audios sofware TYPE: context | IMPORTANCE: 5 TAGS: ['proyecto', 'workspace-root'] Contenido: 10_control_de_plagas.mp4, 11_cosecha.mp4, 12_secado.mp4, 13_trimming.mp4, 14_cuarentena.mp4, 15_fraccionamiento.mp4, 16_almacenamiento.mp4, 1_ingreso_de_insumos.mp4 ====================================================================== TITLE: Proyecto: cannabis-wiki TYPE: context | IMPORTANCE: 5 TAGS: ['proyecto', 'workspace-root'] Eres el mantenedor de este LLM Wiki. Sigue estas reglas. - index.md → Índice principal (SIEMPRE navegar desde acá) - aw/ → Fuentes crudas sin procesar ====================================================================== TITLE: Proyecto: CannTrace Capturas TYPE: context | IMPORTANCE: 5 TAGS: ['proyecto', 'workspace-root'] Contenido: Captura1.PNG, Captura2.PNG, Captura3.PNG, Captura4.PNG, Captura5.PNG, Captura6.PNG ====================================================================== TITLE: Proyecto: codex-trading-plugin TYPE: context | IMPORTANCE: 5 TAGS: ['proyecto', 'workspace-root'] Contenido: .codex-plugin, .mcp.json, skills ====================================================================== TITLE: Proyecto: consultora-ojeda-araujo TYPE: context | IMPORTANCE: 5 TAGS: ['proyecto', 'workspace-root'] Contenido: admin-static.html, consultora-ojeda-araujo.html, img, index-localstorage.html, index-static.html, index.html, index_backup.html, index_vieja.html ====================================================================== TITLE: Proyecto: Gamp5 TYPE: context | IMPORTANCE: 5 TAGS: ['proyecto', 'workspace-root'] Contenido: Checklist _ CUMCS IMC (1).xlsx, CM-RE-1010 Matriz Consolidada v2.xlsx, Disposicion 4159_ Guia BPF anexo_6892007_1 (1).pdf, Documentos para Gastón.xlsx, WhatsApp Ptt 2026-04-16 at 9.23.33 AM.ogg, WhatsApp Ptt 2026-04-17 at 11.50.24 AM.ogg ====================================================================== TITLE: Proyecto: gaston-trading TYPE: context | IMPORTANCE: 5 TAGS: ['proyecto', 'workspace-root'] Sistema multi-activo de trading con senales + analisis automatico + backtest. Pensado para correr en la PC con RTX 5090 (ejecucion + IA local) y desarrollar/probar en esta PC (sin GPU). ``` gaston-trading/ ├── src/ ====================================================================== TITLE: Proyecto: Logos PDF TYPE: context | IMPORTANCE: 5 TAGS: ['proyecto', 'workspace-root'] Contenido: CannTrace — Landing (Print).pdf, CannTrace — Logos (Print).pdf ====================================================================== TITLE: Proyecto: PAQUETE-CONSULTORIA TYPE: context | IMPORTANCE: 5 TAGS: ['proyecto', 'workspace-root'] Contenido: consultora-ojeda-araujo, datos, INFORME_FINAL_PLATAFORMA.md, INFORME_FINAL_PLATAFORMA.pdf, INSTALAR_CONSULTORIA.bat, INSTRUCCIONES_PARA_IA.md, LEEME_PRIMERO.txt, plataforma-cursos ====================================================================== TITLE: Proyecto: plataforma-cursos TYPE: context | IMPORTANCE: 5 TAGS: ['proyecto', 'workspace-root'] Contenido: Trasabilidad ====================================================================== TITLE: Proyecto: plugins TYPE: context | IMPORTANCE: 5 TAGS: ['proyecto', 'workspace-root'] Contenido: gaston-trading ====================================================================== TITLE: Proyecto: temp_imgs TYPE: context | IMPORTANCE: 5 TAGS: ['proyecto', 'workspace-root'] Contenido: img_26027.jpg, img_26028.jpg, img_26029.jpg, img_26030.jpg, img_26031.jpg, img_26032.jpg, img_26033.jpg, screenshot1.png ====================================================================== TITLE: Proyecto: trabajos-natalia TYPE: context | IMPORTANCE: 5 TAGS: ['proyecto', 'workspace-root'] Contenido: Araujo y Ojeda & Asociados _ Consultoría en Agro, Regulación y Sustentabilidad.html, Araujo y Ojeda & Asociados _ Consultoría en Agro, Regulación y Sustentabilidad_files, index.html, Landing Page Araujo y Ojeda & Asociados.html, Landing Page Araujo y Ojeda & Asociados_files, Landing page.rar, ojeda-araujo-final.html, ojeda-araujo-final2.html ====================================================================== TITLE: Proyecto: TradingViewExtracted TYPE: context | IMPORTANCE: 5 TAGS: ['proyecto', 'workspace-root'] Contenido: Contents, TradingView.zip ====================================================================== TITLE: AI Stack Docker Gaston — arquitectura completa 2026-04-24 TYPE: infra | IMPORTANCE: 5 TAGS: [] Stack AI casero de Gastón en Docker Desktop Windows. Ubicación: F:/gaston-workspace/ai-stack/docker-compose.yml **4 containers activos:** - openwebui :3000 (v0.9.2, admin gastonrpersoglia@gmail.com) - lightrag :9622 (v1.4.15, workspace=gaston) - mcpo :8770 (MCPO proxy MCPs a OpenWebUI) - hermes :8642 (custom image gaston/hermes-shim:latest — Hermes Agent + FastAPI shim OpenAI-compat) **MCPO-Windows (host):** :8771 corriendo via F:/gaston-workspace/ai-stack/start-mcpo-windows.ps1 con 7 MCPs Windows-only (tradingview-atila, tradingview-cdp, desktop-commander, excel, token-savior, telegram, ruflo). **Red Docker:** ai-net (compartida entre containers). **Volúmenes named:** openwebui_data, hermes_data, lightrag_storage (mapeado a F:/gaston-workspace/gaston-trading/tools/LightRAG/rag_storage). Para poder reanudar en nueva sesión sin re-descubrir la arquitectura Leer docker-compose.yml en ai-stack/, docker ps, y los archivos de configuración de cada componente 9 ["ai-stack", "docker", "openwebui", "lightrag", "hermes", "ruflo", "mcpo"] F:\gaston-workspace ====================================================================== TITLE: CannTrace Supabase YA implementado — proyecto sqdqvhjlmdweuuncwlfb TYPE: infra | IMPORTANCE: 5 TAGS: [] Proyecto Supabase 'CannTrace' (sqdqvhjlmdweuuncwlfb, ACTIVE_HEALTHY, sa-east-1, Postgres 17) tiene 38 tablas con RLS, ~10K filas reales, 47 migrations, 3 Edge Functions (agent-cumcs v11, ai-extract-operacion v15, ai-next-question v8). La matriz CM-RE-1010 del xlsx YA está cargada en tipos_registro_cumcs (84 codigos, 10 grupos, schema con codigo+nombre+grupo+procedimiento_referencia+obligatorio+medio_archivo+conservacion+requiere_firma+referencia_cumcs+tabla_destino). Datos operacionales reales: registros_condiciones_ambientales (1104), registros_fitosanitarios (1664), registros_personal (1573), registros_mantenimiento (356), registros_fertilizantes (161), registros_agua (130), registros_calidad (128), registros_trazabilidad (70), registros_cosecha (15), registros_documentales (12). Audit log: 5190 entradas. Alertas operativas auto-generadas: 665. Lotes: 105. Operaciones: 9. Variedades: 3. Instalaciones: 11. Productos: 8. SOPs: 4. Chat sessions agente: 1. Pending inserts (drafts TTL 10min): 1. Extensions habilitadas: pgcrypto, uuid-ossp, citext, plpgsql, pg_stat_statements, pg_graphql, supabase_vault. Disponibles para activar: vector (pgvector), pgroonga, pg_trgm, pg_cron, pg_net, pgmq, postgis. RPCs implementadas: arbol_genealogico, mapa_instalaciones, COA_digital, audit_trail_lote, alertas_trazabilidad, metricas, sankey, busqueda_event_sourcing, reprocann_agregado. Firmas electrónicas con PNG (compliance 21 CFR Part 11). Realtime activo en CUMCS tables. Lote padre matching automático por substring. Para que cualquier Claude Code futuro sepa instantáneamente lo que YA está en Supabase y NO duplique trabajo (e.g. parsear xlsx que ya están en BD). Es la fuente de verdad estructurada del proyecto, mejor que LightRAG/RAG para queries específicas. Cuando el usuario hable de CannTrace, registros CM-RE, trazabilidad cannabis o quiera leer/modificar datos: usar mcp__supabase__execute_sql en project_id=sqdqvhjlmdweuuncwlfb. NO parsear xlsx ni usar LightRAG para esto. SQL puro vía Supabase MCP. Tabla maestra: tipos_registro_cumcs. Datos operacionales: registros_ con tabla_destino mapeado. 10 ["canntrace", "supabase", "cumcs", "matriz-cm-re", "infra-real", "no-parsear-xlsx", "sql-via-mcp"] F:\gaston-workspace ["Proyecto Supabase ID: sqdqvhjlmdweuuncwlfb", "Org ID: ergjoigiuyptaqdegyak (OA Consultora)", "84 codigos CM-RE en 10 grupos", "5500+ filas operacionales reales", "47 migrations aplicadas Apr 2026", "3 Edge Functions activas", "RLS habilitado en TODAS las tablas", "audit_log con 5190 entradas", "Postgres 17.6.1.104 en sa-east-1", "Segundo proyecto Oa Consultora INACTIVE id=zhdcwcpitegxmewxvqcl"] ====================================================================== TITLE: README TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:agents', 'doc-subproyecto'] [agents/TradingAgents/README.md]

arXiv Discord
LightRAG Logo
HKUDS%2FLightRAG | Trendshift Software packages requiring `transformers`, `torch`, or `cuda` will not be included in the offline depen ====================================================================== TITLE: Programming With LightRAG Core TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/LightRAG/docs/ProgramingWithCore.md] > If you want to integrate LightRAG into your project, we recommend using the REST API provided by the LightRAG Server. LightRAG Core is intended for embedded applications or researchers conducting studies and evaluations. ```python import os import asyncio from lightrag import LightRAG, QueryParam from lightrag.llm.openai import gpt_4o_mini_complete, gpt_4o_complete, openai_embed ====================================================================== TITLE: Evaluation Result Reproduce TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/LightRAG/docs/Reproduce.md] The dataset used in LightRAG can be downloaded from [TommyChien/UltraDomain](https://huggingface.co/datasets/TommyChien/UltraDomain). LightRAG uses the following prompt to generate high-level queries, with the corresponding code in `examples/generate_query.py`. **Prompt** ``` Given the following description of a dataset: ====================================================================== TITLE: uv.lock Update Guide TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/LightRAG/docs/UV_LOCK_GUIDE.md] `uv.lock` is uv's lock file. It captures the exact version of every dependency, including transitive ones, much like: - Node.js `package-lock.json` - Rust `Cargo.lock` - Python Poetry `poetry.lock` Keeping `uv.lock` in version control guarantees that everyone installs the same dependency set. - Running `uv sync --frozen` ====================================================================== TITLE: LightRAG Helm Chart TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/LightRAG/k8s-deploy/README.md] This is the Helm chart for LightRAG, used to deploy LightRAG services on a Kubernetes cluster. There are two recommended deployment methods for LightRAG: 1. **Lightweight Deployment**: Using built-in lightweight storage, suitable for testing and small-scale usage 2. **Production Deployment**: Using external databases (such as PostgreSQL and Neo4J), suitable for production environments and large-sc ====================================================================== TITLE: LightRAG WebUI TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/LightRAG/lightrag_webui/README.md] LightRAG WebUI is a React-based web interface for interacting with the LightRAG system. It provides a user-friendly interface for querying, managing, and exploring LightRAG's functionalities. 1. **Install Bun:** If you haven't already installed Bun, follow the official documentation: [https://bun.sh/docs/installation](https://bun.sh/docs/installation) ====================================================================== TITLE: LightRAG 3D Graph Viewer TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/LightRAG/lightrag/tools/lightrag_visualizer/README.md] An interactive 3D graph visualization tool included in the LightRAG package for visualizing and analyzing RAG (Retrieval-Augmented Generation) graphs and other graph structures. ![image](https://github.com/user-attachments/assets/b0d86184-99fc-468c-96ed-c611f14292bf) ```bash pip install lightrag-hku[tools] # Install with visualization tool only ====================================================================== TITLE: Sample Documents for Evaluation TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/LightRAG/lightrag/evaluation/sample_documents/README.md] These markdown files correspond to test questions in `../sample_dataset.json`. 1. **Index documents** into LightRAG (via WebUI, API, or Python) 2. **Run evaluation**: `python lightrag/evaluation/eval_rag_quality.py` 3. **Expected results**: ~91-100% RAGAS score per question - `01_lightrag_overview.md` - LightRAG framework and hallucination problem ====================================================================== TITLE: Using KubeBlocks to Deploy and Manage Databases TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/LightRAG/k8s-deploy/databases/README.md] Learn how to quickly deploy and manage various databases in a Kubernetes (K8s) environment through KubeBlocks. KubeBlocks is a production-ready, open-source toolkit that runs any database--SQL, NoSQL, vector, or document--on Kubernetes. It scales smoothly from quick dev tests to full production clusters, making it a solid choice for RAG workloads like FastGPT that need several data stores working ====================================================================== TITLE: Everything Claude Code for CodeBuddy TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/.codebuddy/README.md] Bring Everything Claude Code (ECC) workflows to CodeBuddy IDE. This repository provides custom commands, agents, skills, and rules that can be installed into any CodeBuddy project using the unified Target Adapter architecture. Use the unified install system for full lifecycle management: ```bash node scripts/install-apply.js --target codebuddy --profile developer ====================================================================== TITLE: OpenCode ECC Plugin TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/.opencode/README.md] > WARNING: This README is specific to OpenCode usage. > If you installed ECC via npm (e.g. `npm install opencode-ecc`), refer to the root README instead. Everything Claude Code (ECC) plugin for OpenCode - agents, commands, hooks, and skills. There are two ways to use Everything Claude Code (ECC): 1. **npm package (recommended for most users)** ====================================================================== TITLE: Architecture Improvement Recommendations TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/ARCHITECTURE-IMPROVEMENTS.md] This document captures architect-level improvements for the Everything Claude Code (ECC) project. It is written from the perspective of a Claude Code coding architect aiming to improve maintainability, consistency, and long-term quality. --- **Issue:** AGENTS.md states "13 specialized agents, 50+ skills, 33 commands" while the repo has **16 agents**, **65+ skills**, and **40 commands**. README and ====================================================================== TITLE: Continuous Learning v2 Spec TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/continuous-learning-v2-spec.md] This document captures the v2 continuous-learning architecture: 1. Hook-based observation capture 2. Background observer analysis loop 3. Instinct scoring and persistence 4. Evolution of instincts into reusable skills/commands Primary implementation lives in: - `skills/continuous-learning-v2/` - `scripts/hooks/` ====================================================================== TITLE: Hermes / OpenClaw -> ECC Migration TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/HERMES-OPENCLAW-MIGRATION.md] This document is the public migration guide for moving a Hermes or OpenClaw-style operator setup into the current ECC model. The goal is not to reproduce a private operator workspace byte-for-byte. The goal is to preserve the useful workflow surface: - reusable skills - stable automation entrypoints - cross-harness portability ====================================================================== TITLE: Hermes x ECC Setup TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/HERMES-SETUP.md] Hermes is the operator shell. ECC is the reusable system behind it. This guide is the public, sanitized version of the Hermes stack used to run content, outreach, research, sales ops, finance checks, and engineering workflows from one terminal-native surface. - ECC skills, agents, commands, hooks, and MCP configs from this repo ====================================================================== TITLE: Mega Plan Repo Prompt List — March 12, 2026 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/MEGA-PLAN-REPO-PROMPTS-2026-03-12.md] Use these prompts to split the remaining March 11 mega-plan work by repo. They are written for parallel agents and assume the March 12 orchestration and Windows CI lane is already merged via `#417`. - `everything-claude-code` has finished the orchestration, Codex baseline, and Windows CI recovery lane. - The next open ECC Phase 1 items are: ====================================================================== TITLE: Example Project CLAUDE.md TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/examples/CLAUDE.md] This is an example project-level CLAUDE.md file. Place this in your project root. [Brief description of your project - what it does, tech stack] - Many small files over few large files - High cohesion, low coupling - 200-400 lines typical, 800 max per file - Organize by feature/domain, not by type - No emojis in code, comments, or documentation ====================================================================== TITLE: Metrics and Sponsorship Playbook TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/business/metrics-and-sponsorship.md] This file is a practical script for sponsor calls and ecosystem partner reviews. Use four categories in every update: 1. **Distribution** — npm packages and GitHub App installs 2. **Adoption** — stars, forks, contributors, release cadence 3. **Product surface** — commands/skills/agents and cross-platform support ====================================================================== TITLE: Product Capability Template TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/examples/product-capability-template.md] Use this when product intent exists but the implementation constraints are still implicit. The purpose is to create a durable capability contract, not another vague planning doc. - **Capability name:** - **Source:** PRD / issue / discussion / roadmap / founder note - **Primary actor:** - **Outcome after ship:** ====================================================================== TITLE: Contribuindo para o Everything Claude Code TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/CONTRIBUTING.md] Obrigado por querer contribuir! Este repositório é um recurso comunitário para usuários do Claude Code. - [O Que Estamos Buscando](#o-que-estamos-buscando) - [Início Rápido](#início-rápido) - [Contribuindo com Skills](#contribuindo-com-skills) - [Contribuindo com Agentes](#contribuindo-com-agentes) - [Contribuindo com Hooks](#contribuindo-com-hooks) ====================================================================== TITLE: Glossário de Terminologia (TERMINOLOGY) TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/TERMINOLOGY.md] Este documento registra a correspondência de termos utilizados nas traduções para português brasileiro (pt-BR), garantindo consistência. - **Confirmado**: Tradução confirmada - **Pendente**: Aguardando revisão --- | English | pt-BR | Status | Observações | |---------|-------|--------|-------------| | Agent | Agent | Confirmado | Manter em inglês | ====================================================================== TITLE: Everything Claude Code (ECC) — Agent Talimatları TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/AGENTS.md] Bu, yazılım geliştirme için 28 özel agent, 116 skill, 59 command ve otomatik hook iş akışları sağlayan **üretime hazır bir AI kodlama eklentisidir**. **Sürüm:** 1.10.0 1. **Agent-Öncelikli** — Alan görevleri için özel agentlara delege edin 2. **Test-Odaklı** — Uygulamadan önce testler yazın, %80+ kapsama gereklidir ====================================================================== TITLE: Katkıda Bulunanlar Sözleşmesi Davranış Kuralları TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/CODE_OF_CONDUCT.md] Üyeler, katkıda bulunanlar ve liderler olarak, topluluğumuza katılımı yaş, beden ölçüsü, görünür veya görünmez engellilik, etnik köken, cinsiyet özellikleri, cinsiyet kimliği ve ifadesi, deneyim seviyesi, eğitim, sosyo-ekonomik durum, milliyet, kişisel görünüm, ırk, din veya cinsel kimlik ve yönelim fark etmeksizin herkes için tacizden arınmış bir deneyim haline getirmeyi taahhüt ediyoruz. ====================================================================== TITLE: Everything Claude Code'a Katkıda Bulunma TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/CONTRIBUTING.md] Katkıda bulunmak istediğiniz için teşekkürler! Bu repo, Claude Code kullanıcıları için bir topluluk kaynağıdır. - [Ne Arıyoruz](#ne-arıyoruz) - [Hızlı Başlangıç](#hızlı-başlangıç) - [Skill'lere Katkıda Bulunma](#skilllere-katkıda-bulunma) - [Agent'lara Katkıda Bulunma](#agentlara-katkıda-bulunma) - [Hook'lara Katkıda Bulunma](#hooklara-katkıda-bulunma) ====================================================================== TITLE: ECC'ye Sponsor Olma TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/SPONSORING.md] ECC, Claude Code, Cursor, OpenCode ve Codex app/CLI genelinde açık kaynaklı bir ajan performans sistemi olarak sürdürülmektedir. Sponsorluk doğrudan şunları destekler: - Daha hızlı hata düzeltme ve sürüm döngüleri - Harness'lar arasında platformlar arası eşitlik çalışması - Topluluk için ücretsiz kalan genel dokümantasyon, beceriler ve güvenilirlik araçları ====================================================================== TITLE: Sponsorlar TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/SPONSORS.md] Bu projeye sponsor olan herkese teşekkürler! Desteğiniz ECC ekosisteminin büyümesini sağlıyor. *Burada yer almak için [Kurumsal sponsor](https://github.com/sponsors/affaan-m) olun* *Burada yer almak için [İşletme sponsoru](https://github.com/sponsors/affaan-m) olun* *Burada yer almak için [Takım sponsoru](https://github.com/sponsors/affaan-m) olun* ====================================================================== TITLE: Terminoloji Tablosu (Terminology Glossary) TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/TERMINOLOGY.md] Bu doküman Türkçe çevirilerin terminoloji karşılıklarını kayıt altına alarak çeviri tutarlılığını sağlar. - **Onaylandı (Confirmed)**: Onaylanmış çeviri - **Beklemede (Pending)**: İnceleme bekleyen çeviri --- | English | tr | Durum | Notlar | |---------|-------|------|------| | Agent | Agent | Onaylandı | İngilizce tutulur | ====================================================================== TITLE: Her Şey Agentic Güvenliğe Dair Kısa Kılavuz TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/the-security-guide.md] _everything claude code / araştırma / güvenlik_ --- Son makalemden bu yana epey zaman geçti. ECC devtooling ekosistemini geliştirmeye zaman harcadım. Bu süreçte sıcak ancak önemli konulardan biri agent güvenliği oldu. Açık kaynak agent'ların yaygın olarak benimsenmesi burada. OpenClaw ve diğerleri bilgisayarınızda dolaşıyor. Claude Code ve Codex (ECC kullanarak) gibi sürekli çalışma harness'leri y ====================================================================== TITLE: Claude Code'un Her Şeyine Dair Kısa Kılavuz TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/the-shortform-guide.md] ![Header: Anthropic Hackathon Winner - Tips & Tricks for Claude Code](../assets/images/shortform/00-header.png) --- **Şubat ayında deneysel kullanıma sunulduğundan beri hevesli bir Claude Code kullanıcısıyım ve [@DRodriguezFX](https://x.com/DRodriguezFX) ile birlikte tamamen Claude Code kullanarak [zenith.chat](https://zenith.chat) projesiyle Anthropic x Forum Ventures hackathon'unu kazandım.** ====================================================================== TITLE: Claude Code 简明指南 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/the-shortform-guide.md] ![标题：Anthropic 黑客马拉松获胜者 - Claude Code 技巧与窍门](../../assets/images/shortform/00-header.png) *** **自 2 月实验性推出以来，我一直是 Claude Code 的忠实用户，并凭借 [zenith.chat](https://zenith.chat) 与 [@DRodriguezFX](https://x.com/DRodriguezFX) 一起赢得了 Anthropic x Forum Ventures 的黑客马拉松——完全使用 Claude Code。** 经过 10 个月的日常使用，以下是我的完整设置：技能、钩子、子代理、MCP、插件以及实际有效的方法。 ====================================================================== TITLE: 安全性審查員 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-TW/agents/security-reviewer.md] 您是一位專注於識別和修復 Web 應用程式弱點的安全性專家。您的任務是透過對程式碼、設定和相依性進行徹底的安全性審查，在問題進入生產環境之前預防安全性問題。 1. **弱點偵測** - 識別 OWASP Top 10 和常見安全性問題 2. **密鑰偵測** - 找出寫死的 API 金鑰、密碼、Token 3. **輸入驗證** - 確保所有使用者輸入都正確清理 4. **驗證/授權** - 驗證適當的存取控制 5. **相依性安全性** - 檢查有弱點的 npm 套件 6. **安全性最佳實務** - 強制執行安全編碼模式 - **npm audit** - 檢查有弱點的相依性 - **eslint-plugin-security** - 安全性問題的靜態分析 ====================================================================== TITLE: 常見模式 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-TW/rules/patterns.md] ```typescript interface ApiResponse { success: boolean data?: T error?: string meta?: { total: number page: number limit: number } } ``` ```typescript export function useDebounce(value: T, delay: number): T { const [debouncedValue, setDebouncedValue] = useState(value) useEffect(() => { const handler = setTimeout(() => setDebouncedValue(value), delay) ====================================================================== TITLE: 前端開發模式 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-TW/skills/frontend-patterns/SKILL.md] 用於 React、Next.js 和高效能使用者介面的現代前端模式。 ```typescript // PASS: 良好：元件組合 interface CardProps { children: React.ReactNode variant?: 'default' | 'outlined' } export function Card({ children, variant = 'default' }: CardProps) { return
{children}
} export function CardHeader({ children }: { children: React.ReactNode }) { ====================================================================== TITLE: cloud-infrastructure-security TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-TW/skills/security-review/cloud-infrastructure-security.md] | name | description | |------|-------------| | cloud-infrastructure-security | Use this skill when deploying to cloud platforms, configuring infrastructure, managing IAM policies, setting up logging/monitoring, or implementing CI/CD pipelines. Provides cloud security checklist aligned with best practices. | ====================================================================== TITLE: PyTorch 构建/运行时错误解决器 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/agents/pytorch-build-resolver.md] 你是一名专业的 PyTorch 错误解决专家。你的任务是以**最小、精准的改动**修复 PyTorch 运行时错误、CUDA 问题、张量形状不匹配和训练失败。 1. 诊断 PyTorch 运行时和 CUDA 错误 2. 修复模型各层间的张量形状不匹配 3. 解决设备放置问题（CPU/GPU） 4. 调试梯度计算失败 5. 修复 DataLoader 和数据流水线错误 6. 处理混合精度（AMP）问题 按顺序运行这些命令： ```bash python -c "import torch; print(f'PyTorch: {torch.__version__}, CUDA: {torch.cuda.is_available()}, Device: {torch.cuda.get_device_name(0) if torch.cuda.is_available() else \"CPU\"} ====================================================================== TITLE: 上下文预算优化器 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/commands/context-budget.md] 分析您的 Claude Code 设置中的上下文窗口消耗，并提供可操作的建议以减少令牌开销。 ``` /context-budget [--verbose] ``` * 默认：提供摘要及主要建议 * `--verbose`：按组件提供完整细分 $ARGUMENTS 运行 **context-budget** 技能（`skills/context-budget/SKILL.md`），并输入以下内容： 1. 如果 `$ARGUMENTS` 中存在 `--verbose` 标志，则传递该标志 2. 除非用户另行指定，否则假设为 200K 上下文窗口（Claude Sonnet 默认值） 3. 遵循技能的四个阶段：清单 → 分类 → 检测问题 → 报告 ====================================================================== TITLE: 上下文预算 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/skills/context-budget/SKILL.md] 分析 Claude Code 会话中每个已加载组件的令牌开销，并提供可操作的优化建议以回收上下文空间。 * 会话性能感觉迟缓或输出质量下降 * 你最近添加了许多技能、代理或 MCP 服务器 * 你想知道实际有多少上下文余量 * 计划添加更多组件，需要知道是否有空间 * 运行 `/context-budget` 命令（本技能为其提供支持） 扫描所有组件目录并估算令牌消耗： **代理** (`agents/*.md`) * 统计每个文件的行数和令牌数（单词数 × 1.3） * 提取 `description` 前言长度 * 标记：文件 >200 行（繁重），描述 >30 词（臃肿的前言） **技能** (`skills/*/SKILL.md`) ====================================================================== TITLE: 持续代理循环 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/skills/continuous-agent-loop/SKILL.md] 这是 v1.8+ 的规范循环技能名称。它在保持一个发布版本的兼容性的同时，取代了 `autonomous-loops`。 ```text Start | +-- 需要严格的 CI/PR 控制？ -- yes --> continuous-pr | +-- 需要 RFC 分解？ -- yes --> rfc-dag | +-- 需要探索性并行生成？ -- yes --> infinite | +-- default --> sequential ``` 推荐的生产栈： 1. RFC 分解 (`ralphinho-rfc-pipeline`) 2. 质量门 (`plankton-code-quality` + `/quality-gate`) ====================================================================== TITLE: Java 编码规范 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/skills/java-coding-standards/SKILL.md] 适用于 Spring Boot 服务中可读、可维护的 Java (17+) 代码的规范。 * 在 Spring Boot 项目中编写或审查 Java 代码时 * 强制执行命名、不可变性或异常处理约定时 * 使用记录类、密封类或模式匹配（Java 17+）时 * 审查 Optional、流或泛型的使用时 * 构建包和项目布局时 * 清晰优于巧妙 * 默认不可变；最小化共享可变状态 * 快速失败并提供有意义的异常 * 一致的命名和包结构 ```java // PASS: Classes/Records: PascalCase public class MarketService {} public record Money(BigDecimal amount, Currency currency) {} ====================================================================== TITLE: Ktor 服务器模式 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/skills/kotlin-ktor-patterns/SKILL.md] 使用 Kotlin 协程构建健壮、可维护的 HTTP 服务器的综合 Ktor 模式。 * 构建 Ktor HTTP 服务器 * 配置 Ktor 插件（Auth、CORS、ContentNegotiation、StatusPages） * 使用 Ktor 实现 REST API * 使用 Koin 设置依赖注入 * 使用 testApplication 编写 Ktor 集成测试 * 在 Ktor 中使用 WebSocket ```text src/main/kotlin/ ├── com/example/ │ ├── Application.kt # 入口点，模块配置 │ ├── plugins/ ====================================================================== TITLE: MCP 服务器模式 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/skills/mcp-server-patterns/SKILL.md] 模型上下文协议（MCP）允许 AI 助手调用工具、读取资源和使用来自服务器的提示。在构建或维护 MCP 服务器时使用此技能。SDK API 会演进；请查阅 Context7（查询文档 "MCP"）或官方 MCP 文档以获取当前的方法名称和签名。 在以下情况时使用：实现新的 MCP 服务器、添加工具或资源、选择 stdio 与 HTTP、升级 SDK，或调试 MCP 注册和传输问题。 * **工具**：模型可以调用的操作（例如搜索、运行命令）。根据 SDK 版本，使用 `registerTool()` 或 `tool()` 注册。 * **资源**：模型可以获取的只读数据（例如文件内容、API 响应）。根据 SDK 版本，使用 `registerResource()` 或 `resource()` 注册。处理程序通常接收一个 `uri` 参数。 ====================================================================== TITLE: PyTorch 开发模式 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/skills/pytorch-patterns/SKILL.md] 构建稳健、高效和可复现深度学习应用的 PyTorch 惯用模式与最佳实践。 * 编写新的 PyTorch 模型或训练脚本时 * 评审深度学习代码时 * 调试训练循环或数据管道时 * 优化 GPU 内存使用或训练速度时 * 设置可复现实验时 始终编写能在 CPU 和 GPU 上运行且不硬编码设备的代码。 ```python device = torch.device("cuda" if torch.cuda.is_available() else "cpu") model = MyModel().to(device) data = data.to(device) model = MyModel().cuda() # Crashes if no GPU ====================================================================== TITLE: 云与基础设施安全技能 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/skills/security-review/cloud-infrastructure-security.md] | name | description | |------|-------------| | cloud-infrastructure-security | 在部署到云平台、配置基础设施、管理IAM策略、设置日志记录/监控或实现CI/CD流水线时使用此技能。提供符合最佳实践的云安全检查清单。 | 此技能确保云基础设施、CI/CD流水线和部署配置遵循安全最佳实践并符合行业标准。 * 将应用程序部署到云平台（AWS、Vercel、Railway、Cloudflare） * 配置IAM角色和权限 * 设置CI/CD流水线 * 实施基础设施即代码（Terraform、CloudFormation） * 配置日志记录和监控 ====================================================================== TITLE: Swift 6.2 可接近的并发 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/skills/swift-concurrency-6-2/SKILL.md] 采用 Swift 6.2 并发模型的模式，其中代码默认在单线程上运行，并发是显式引入的。在无需牺牲性能的情况下消除常见的数据竞争错误。 * 将 Swift 5.x 或 6.0/6.1 项目迁移到 Swift 6.2 * 解决数据竞争安全编译器错误 * 设计基于 MainActor 的应用架构 * 将 CPU 密集型工作卸载到后台线程 * 在 MainActor 隔离的类型上实现协议一致性 * 在 Xcode 26 中启用“可接近的并发”构建设置 在 Swift 6.1 及更早版本中，异步函数可能会被隐式卸载到后台线程，即使在看似安全的代码中也会导致数据竞争错误： ```swift // Swift 6.1: ERROR ====================================================================== TITLE: X API TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/skills/x-api/SKILL.md] 以编程方式与 X（Twitter）交互，用于发布、读取、搜索和分析。 * 用户希望以编程方式发布推文或帖子串 * 从 X 读取时间线、提及或用户数据 * 在 X 上搜索内容、趋势或对话 * 构建 X 集成或机器人 * 分析和参与度跟踪 * 用户提及"发布到 X"、"发推"、"X API"或"Twitter API" 最佳适用场景：读取密集型操作、搜索、公开数据。 ```bash export X_BEARER_TOKEN="your-bearer-token" ``` ```python import os import requests bearer = os.environ["X_BEARER_TOKEN"] headers = {"Authorization": f"Bearer {bearer}"} ====================================================================== TITLE: C# 钩子 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/rules/csharp/hooks.md] > 本文档基于 [common/hooks.md](../common/hooks.md) 扩展了 C# 相关的具体内容。 在 `~/.claude/settings.json` 中配置： * **dotnet format**：自动格式化编辑过的 C# 文件并应用分析器修复 * **dotnet build**：验证编辑后解决方案或项目是否仍能编译 * **dotnet test --no-build**：在行为更改后重新运行最近相关的测试项目 * 在结束涉及广泛 C# 更改的会话前，运行一次最终的 `dotnet build` * 当 `appsettings*.json` 文件被修改时发出警告，以防敏感信息被提交 ====================================================================== TITLE: C# 安全性 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/rules/csharp/security.md] > 本文档在 [common/security.md](../common/security.md) 的基础上补充了 C# 特有的内容。 * 切勿在源代码中硬编码 API 密钥、令牌或连接字符串 * 在本地开发环境中使用环境变量或用户密钥，在生产环境中使用密钥管理器 * 确保 `appsettings.*.json` 中不包含真实的凭证信息 ```csharp // BAD const string ApiKey = "sk-live-123"; // GOOD var apiKey = builder.Configuration["OpenAI:ApiKey"] ?? throw new InvalidOperationException("OpenAI:ApiKey is not configured."); ====================================================================== TITLE: Go 模式 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/rules/golang/patterns.md] > 本文档在 [common/patterns.md](../common/patterns.md) 的基础上扩展了 Go 语言特定的内容。 ```go type Option func(*Server) func WithPort(port int) Option { return func(s *Server) { s.port = port } } func NewServer(opts ...Option) *Server { s := &Server{port: 8080} for _, opt := range opts { opt(s) } return s } ``` 在接口被使用的地方定义它们，而不是在它们被实现的地方。 ====================================================================== TITLE: Java 钩子 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/rules/java/hooks.md] > 本文件在[common/hooks.md](../common/hooks.md)的基础上扩展了Java相关的内容。 在 `~/.claude/settings.json` 中配置： * **google-java-format**：编辑后自动格式化 `.java` 文件 * **checkstyle**：编辑Java文件后运行样式检查 * **./mvnw compile** 或 **./gradlew compileJava**：变更后验证编译 ====================================================================== TITLE: Java 安全 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/zh-CN/rules/java/security.md] > 本文档在 [common/security.md](../common/security.md) 的基础上，补充了 Java 相关的内容。 * 切勿在源代码中硬编码 API 密钥、令牌或凭据 * 使用环境变量：`System.getenv("API_KEY")` * 生产环境密钥请使用密钥管理器（如 Vault、AWS Secrets Manager） * 包含密钥的本地配置文件应放在 `.gitignore` 中 ```java // BAD private static final String API_KEY = "sk-abc123..."; // GOOD — environment variable ====================================================================== TITLE: Rolünüz TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/agents/architect.md] Ölçeklenebilir, sürdürülebilir sistem tasarımında uzmanlaşmış kıdemli bir yazılım mimarısınız. - Yeni özellikler için sistem mimarisi tasarlayın - Teknik ödünleşimleri değerlendirin - Kalıpları ve en iyi uygulamaları önerin - Ölçeklenebilirlik darboğazlarını belirleyin - Gelecekteki büyüme için planlayın - Kod tabanı genelinde tutarlılık sağlayın ====================================================================== TITLE: Documentation & Codemap Specialist TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/agents/doc-updater.md] Codemap'leri ve dokümantasyonu kod tabanıyla güncel tutan bir dokümantasyon specialistisiniz. Misyonunuz, kodun gerçek durumunu yansıtan doğru, güncel dokümantasyon sürdürmektir. 1. **Codemap Oluşturma** — Kod tabanı yapısından mimari haritalar oluşturun 2. **Dokümantasyon Güncellemeleri** — README'leri ve kılavuzları koddan yenileyin ====================================================================== TITLE: E2E Test Runner TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/agents/e2e-runner.md] Bir uzman uçtan uca test specialistisiniz. Misyonunuz, uygun artifact yönetimi ve kararsız test işleme ile kapsamlı E2E testleri oluşturarak, sürdürerek ve çalıştırarak kritik kullanıcı yolculuklarının doğru çalıştığından emin olmaktır. 1. **Test Yolculuğu Oluşturma** — Kullanıcı akışları için testler yazın (Agent Browser tercih edin, Playwright'a geri dönün) ====================================================================== TITLE: Görev TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/agents/harness-optimizer.md] Koşum iyileştiricisisiniz. Ürün kodunu yeniden yazmak yerine koşum yapılandırmasını iyileştirerek agent tamamlama kalitesini artırın. 1. `/harness-audit` çalıştırın ve temel skor toplayın. 2. En önemli 3 kaldıraç alanını belirleyin (kancalar, değerlendirmeler, yönlendirme, bağlam, güvenlik). 3. Minimal, geri alınabilir yapılandırma değişiklikleri önerin. ====================================================================== TITLE: java-reviewer TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/agents/java-reviewer.md] Idiomatic Java ve Spring Boot best practice'lerinin yüksek standartlarını sağlayan kıdemli bir Java mühendisisiniz. Çağrıldığında: 1. Son Java dosya değişikliklerini görmek için `git diff -- '*.java'` çalıştırın 2. Varsa `mvn verify -q` veya `./gradlew check` çalıştırın 3. Değiştirilmiş `.java` dosyalarına odaklanın ====================================================================== TITLE: Kotlin Build Error Resolver TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/agents/kotlin-build-resolver.md] Uzman bir Kotlin/Gradle build hata çözümleme uzmanısınız. Misyonunuz, Kotlin build hatalarını, Gradle konfigürasyon sorunlarını ve dependency çözümleme başarısızlıklarını **minimal, cerrahi değişikliklerle** düzeltmektir. 1. Kotlin derleme hatalarını teşhis etme 2. Gradle build konfigürasyon sorunlarını düzeltme ====================================================================== TITLE: PyTorch Build/Runtime Error Resolver TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/agents/pytorch-build-resolver.md] Uzman bir PyTorch hata çözümleme uzmanısınız. Misyonunuz, PyTorch runtime hatalarını, CUDA sorunlarını, tensor shape uyumsuzluklarını ve training başarısızlıklarını **minimal, cerrahi değişikliklerle** düzeltmektir. 1. PyTorch runtime ve CUDA hatalarını teşhis etme 2. Model katmanları boyunca tensor shape uyumsuzluklarını düzeltme ====================================================================== TITLE: Security Reviewer TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/agents/security-reviewer.md] Web uygulamalarındaki güvenlik açıklarını belirleme ve düzeltmeye odaklanan uzman bir güvenlik specialistisiniz. Misyonunuz, güvenlik sorunlarının production'a ulaşmadan önce önlenmesidir. 1. **Güvenlik Açığı Tespiti** — OWASP Top 10 ve yaygın güvenlik sorunlarını belirleyin 2. **Secret Tespiti** — Sabit kodlanmış API anahtarlarını, parolaları, token'ları bulun ====================================================================== TITLE: Code Review TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/commands/code-review.md] Commit edilmemiş değişikliklerin kapsamlı güvenlik ve kalite incelemesi: 1. Değişen dosyaları al: git diff --name-only HEAD 2. Her değişen dosya için şunları kontrol et: **Güvenlik Sorunları (KRİTİK):** - Hardcode edilmiş kimlik bilgileri, API anahtarları, token'lar - SQL injection açıklıkları - XSS açıklıkları ====================================================================== TITLE: E2E Komutu TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/commands/e2e.md] Bu komut, Playwright kullanarak end-to-end testleri oluşturmak, sürdürmek ve yürütmek için **e2e-runner** agent'ını çağırır. 1. **Test Yolculukları Oluştur** - Kullanıcı akışları için Playwright testleri oluştur 2. **E2E Testlerini Çalıştır** - Testleri tarayıcılar arasında yürüt 3. **Artifact'ları Yakala** - Hatalarda ekran görüntüleri, videolar, izlemeler ====================================================================== TITLE: Go Build and Fix TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/commands/go-build.md] Bu komut, minimal değişikliklerle Go build hatalarını aşamalı olarak düzeltmek için **go-build-resolver** agent'ını çağırır. 1. **Diagnostics Çalıştır**: `go build`, `go vet`, `staticcheck` yürüt 2. **Hataları Parse Et**: Dosyaya göre grupla ve önem derecesine göre sırala 3. **Aşamalı Düzelt**: Bir seferde bir hata ====================================================================== TITLE: Go Code Review TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/commands/go-review.md] Bu komut, Go'ya özel kapsamlı kod incelemesi için **go-reviewer** agent'ını çağırır. 1. **Go Değişikliklerini Tanımla**: `git diff` ile değiştirilmiş `.go` dosyalarını bul 2. **Static Analiz Çalıştır**: `go vet`, `staticcheck` ve `golangci-lint` yürüt 3. **Güvenlik Taraması**: SQL injection, command injection, race condition'ları kontrol et ====================================================================== TITLE: Backend - Backend Odaklı Geliştirme TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/commands/multi-backend.md] Backend odaklı iş akışı (Research → Ideation → Plan → Execute → Optimize → Review), Codex liderliğinde. ```bash /backend ``` - Backend task: $ARGUMENTS - Codex liderliğinde, Gemini yardımcı referans için - Uygulanabilir: API tasarımı, algoritma implementasyonu, veritabanı optimizasyonu, business logic ====================================================================== TITLE: Frontend - Frontend Odaklı Geliştirme TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/commands/multi-frontend.md] Frontend odaklı iş akışı (Research → Ideation → Plan → Execute → Optimize → Review), Gemini liderliğinde. ```bash /frontend ``` - Frontend task: $ARGUMENTS - Gemini liderliğinde, Codex yardımcı referans için - Uygulanabilir: Component tasarımı, responsive layout, UI animasyonları, stil optimizasyonu ====================================================================== TITLE: Sessions Komutu TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/commands/sessions.md] Claude Code session geçmişini yönet - `~/.claude/sessions/` dizininde saklanan session'ları listele, yükle, alias ata ve düzenle. `/sessions [list|load|alias|info|help] [options]` Tüm session'ları metadata, filtreleme ve sayfalama ile göster. Bir swarm için operatör-yüzey context'e ihtiyacınız olduğunda `/sessions info` kullanın: branch, worktree yolu ve session güncelliği. ====================================================================== TITLE: Paket Yöneticisi Kurulumu TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/commands/setup-pm.md] Bu proje veya global olarak tercih ettiğiniz paket yöneticisini yapılandırın. ```bash node scripts/setup-package-manager.js --detect node scripts/setup-package-manager.js --global pnpm node scripts/setup-package-manager.js --project bun node scripts/setup-package-manager.js --list ``` Hangi paket yöneticisinin kullanılacağını belirlerken, şu sıra kontrol edilir: ====================================================================== TITLE: Update Documentation TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/commands/update-docs.md] Dokümanları codebase ile senkronize et, truth-of-source dosyalarından oluştur. | Kaynak | Oluşturur | |--------|-----------| | `package.json` scripts | Mevcut komutlar referansı | | `.env.example` | Environment variable dokümanı | | `openapi.yaml` / route dosyaları | API endpoint referansı | | Kaynak kod export'ları | Public API dokümanı | ====================================================================== TITLE: Geliştirme Bağlamı TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/contexts/dev.md] Mod: Aktif geliştirme Odak: Uygulama, kodlama, özellik geliştirme - Önce kod yaz, sonra açıkla - Mükemmel çözümler yerine çalışan çözümleri tercih et - Değişikliklerden sonra testleri çalıştır - Commit'leri atomik tut 1. Çalışır hale getir 2. Doğru hale getir 3. Temiz hale getir - Kod değişiklikleri için Edit, Write ====================================================================== TITLE: Araştırma Bağlamı TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/contexts/research.md] Mod: Keşif, inceleme, öğrenme Odak: Harekete geçmeden önce anlama - Sonuca varmadan önce geniş kapsamlı oku - Açıklayıcı sorular sor - İlerledikçe bulguları belge - Anlayış netleşene kadar kod yazma 1. Soruyu anla 2. İlgili kod/belgeleri keşfet 3. Hipotez oluştur 4. Kanıtlarla doğrula 5. Bulguları özetle - Kodu anlamak için Read ====================================================================== TITLE: Örnek Proje CLAUDE.md TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/examples/CLAUDE.md] Bu, örnek bir proje seviyesi CLAUDE.md dosyasıdır. Bunu proje kök dizininize yerleştirin. [Projenizin kısa açıklaması - ne yaptığı, teknoloji yığını] - Birkaç büyük dosya yerine çok sayıda küçük dosya - Yüksek bağlılık, düşük bağımlılık - Tipik olarak 200-400 satır, dosya başına maksimum 800 satır - Tipe göre değil, özellik/domain'e göre organize edin ====================================================================== TITLE: Backend Geliştirme Kalıpları TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/skills/backend-patterns/SKILL.md] Ölçeklenebilir sunucu tarafı uygulamalar için backend mimari kalıpları ve en iyi uygulamalar. - REST veya GraphQL API endpoint'leri tasarlarken - Repository, service veya controller katmanları uygularken - Veritabanı sorgularını optimize ederken (N+1, indeksleme, bağlantı havuzu) - Önbellekleme eklerken (Redis, in-memory, HTTP cache başlıkları) ====================================================================== TITLE: Deployment Kalıpları TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/skills/deployment-patterns/SKILL.md] Üretim deployment iş akışları ve CI/CD en iyi uygulamaları. - CI/CD pipeline'ları kurarken - Bir uygulamayı Docker'ize ederken - Deployment stratejisi planlarken (blue-green, canary, rolling) - Sağlık kontrolleri ve hazırlık probe'ları uygularken - Üretim yayınına hazırlanırken - Ortama özgü ayarları yapılandırırken ====================================================================== TITLE: Django Geliştirme Desenleri TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/skills/django-patterns/SKILL.md] Ölçeklenebilir, bakımı kolay uygulamalar için production-grade Django mimari desenleri. - Django web uygulamaları oluştururken - Django REST Framework API'leri tasarlarken - Django ORM ve modeller ile çalışırken - Django proje yapısını kurarken - Caching, signal'ler, middleware implement ederken ``` myproject/ ====================================================================== TITLE: Docker Kalıpları TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/skills/docker-patterns/SKILL.md] Konteynerize edilmiş geliştirme için Docker ve Docker Compose en iyi uygulamaları. - Yerel geliştirme için Docker Compose kurarken - Çok konteynerli mimariler tasarlarken - Konteyner ağ veya volume sorunlarını giderirken - Dockerfile'ları güvenlik ve boyut için incelerken - Yerel geliştirmeden konteynerize iş akışına geçerken ====================================================================== TITLE: Eval Harness Skill TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/skills/eval-harness/SKILL.md] Claude Code oturumları için eval-driven development (EDD) ilkelerini uygulayan formal değerlendirme çerçevesi. - AI destekli iş akışları için eval-driven development (EDD) kurarken - Claude Code görev tamamlama için geçti/kaldı kriterleri tanımlarken - pass@k metrikleriyle agent güvenilirliğini ölçerken ====================================================================== TITLE: Frontend Geliştirme Kalıpları TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/skills/frontend-patterns/SKILL.md] React, Next.js ve performanslı kullanıcı arayüzleri için modern frontend kalıpları. - React bileşenleri oluştururken (composition, props, rendering) - State yönetirken (useState, useReducer, Zustand, Context) - Veri çekme implementasyonu (SWR, React Query, server components) - Performans optimize ederken (memoization, virtualization, code splitting) ====================================================================== TITLE: JPA/Hibernate Kalıpları TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/skills/jpa-patterns/SKILL.md] Spring Boot'ta veri modelleme, repository'ler ve performans ayarlaması için kullanın. - JPA entity'leri ve tablo eşlemelerini tasarlarken - İlişkileri tanımlarken (@OneToMany, @ManyToOne, @ManyToMany) - Sorguları optimize ederken (N+1 önleme, fetch stratejileri, projections) - Transaction'ları, auditing'i veya soft delete'leri yapılandırırken ====================================================================== TITLE: Kotlin Geliştirme Kalıpları TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/skills/kotlin-patterns/SKILL.md] Sağlam, verimli ve sürdürülebilir uygulamalar oluşturmak için idiomatic Kotlin kalıpları ve en iyi uygulamalar. - Yeni Kotlin kodu yazarken - Kotlin kodunu incelerken - Mevcut Kotlin kodunu refactor ederken - Kotlin modülleri veya kütüphaneleri tasarlarken - Gradle Kotlin DSL build'lerini yapılandırırken Bu skill yedi temel alanda idiomatic Kotlin konvansiyonlarını uygular: tip sistemi ve safe-cal ====================================================================== TITLE: Laravel Geliştirme Desenleri TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/skills/laravel-patterns/SKILL.md] Ölçeklenebilir, bakım yapılabilir uygulamalar için üretim seviyesi Laravel mimari desenleri. - Laravel web uygulamaları veya API'ler oluşturma - Controller'lar, servisler ve domain mantığını yapılandırma - Eloquent model'ler ve ilişkiler ile çalışma - Resource'lar ve sayfalama ile API tasarlama - Kuyruklar, event'ler, caching ve arka plan işleri ekleme ====================================================================== TITLE: Laravel TDD İş Akışı TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/skills/laravel-tdd/SKILL.md] 80%+ kapsam (unit + feature) ile Laravel uygulamaları için test-driven development. - Laravel'de yeni özellikler veya endpoint'ler - Bug düzeltmeleri veya refactoring'ler - Eloquent model'leri, policy'leri, job'ları ve notification'ları test etme - Proje zaten PHPUnit'te standartlaşmamışsa yeni testler için Pest'i tercih edin ====================================================================== TITLE: Güvenlik İnceleme Skill'i TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/skills/security-review/SKILL.md] Bu skill tüm kodun güvenlik en iyi uygulamalarını takip etmesini sağlar ve potansiyel güvenlik açıklarını tanımlar. - Kimlik doğrulama veya yetkilendirme uygularken - Kullanıcı girdisi veya dosya yüklemeleri işlerken - Yeni API endpoint'leri oluştururken - Secret'lar veya kimlik bilgileriyle çalışırken - Ödeme özellikleri uygularken ====================================================================== TITLE: Spring Boot Güvenlik İncelemesi TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/skills/springboot-security/SKILL.md] Auth ekleme, girişi işleme, endpoint oluşturma veya gizli bilgilerle uğraşırken kullanın. - Kimlik doğrulama ekleme (JWT, OAuth2, session-based) - Yetkilendirme uygulama (@PreAuthorize, role-based erişim) - Kullanıcı girişini doğrulama (Bean Validation, custom validator'lar) - CORS, CSRF veya güvenlik başlıklarını yapılandırma ====================================================================== TITLE: Spring Boot TDD İş Akışı TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/skills/springboot-tdd/SKILL.md] 80%+ kapsam (unit + integration) ile Spring Boot servisleri için TDD rehberi. - Yeni özellikler veya endpoint'ler - Bug düzeltmeleri veya refactoring'ler - Veri erişim mantığı veya güvenlik kuralları ekleme 1) Önce testleri yazın (başarısız olmalılar) 2) Geçmek için minimal kod uygulayın 3) Testleri yeşil tutarken refactor edin ====================================================================== TITLE: Hooks Sistemi TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/rules/common/hooks.md] - **PreToolUse**: Tool yürütmeden önce (validasyon, parametre değişikliği) - **PostToolUse**: Tool yürütmeden sonra (auto-format, kontroller) - **Stop**: Session bittiğinde (final doğrulama) Dikkatli kullan: - Güvenilir, iyi tanımlanmış planlar için etkinleştir - Keşifsel çalışmalar için devre dışı bırak - Asla dangerously-skip-permissions flag'i kullanma ====================================================================== TITLE: Performans Optimizasyonu TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/rules/common/performance.md] **Haiku 4.5** (Sonnet kapasitesinin %90'ı, 3x maliyet tasarrufu): - Sık çağrılan hafif agent'lar - Pair programming ve kod üretimi - Multi-agent sistemlerinde worker agent'lar **Sonnet 4.6** (En iyi kodlama modeli): - Ana geliştirme çalışması - Multi-agent iş akışlarını orkestrasyon - Karmaşık kodlama görevleri ====================================================================== TITLE: Go Pattern'leri TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/rules/golang/patterns.md] > Bu dosya [common/patterns.md](../common/patterns.md) dosyasını Go'ya özgü içerikle genişletir. ```go type Option func(*Server) func WithPort(port int) Option { return func(s *Server) { s.port = port } } func NewServer(opts ...Option) *Server { s := &Server{port: 8080} for _, opt := range opts { opt(s) } return s ====================================================================== TITLE: Python Hooks TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/rules/python/hooks.md] > Bu dosya [common/hooks.md](../common/hooks.md) dosyasını Python'a özgü içerikle genişletir. `~/.claude/settings.json` içinde yapılandır: - **black/ruff**: Edit'ten sonra `.py` dosyalarını otomatik formatla - **mypy/pyright**: `.py` dosyalarını düzenledikten sonra tip kontrolü çalıştır - Düzenlenen dosyalarda `print()` ifadeleri hakkında uyar (bunun yerine `logging` modülü kullan) ====================================================================== TITLE: TypeScript/JavaScript Kodlama Stili TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/rules/typescript/coding-style.md] > Bu dosya [common/coding-style.md](../common/coding-style.md) dosyasını TypeScript/JavaScript'e özgü içerikle genişletir. Public API'ları, paylaşılan modelleri ve component prop'larını açık, okunabilir ve yeniden kullanılabilir yapmak için tipleri kullan. - Dışa aktarılan fonksiyonlara, paylaşılan utility'lere ve public sınıf metotlarına parametre ve dönüş tipleri ekle ====================================================================== TITLE: TypeScript/JavaScript Hooks TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/rules/typescript/hooks.md] > Bu dosya [common/hooks.md](../common/hooks.md) dosyasını TypeScript/JavaScript'e özgü içerikle genişletir. `~/.claude/settings.json` içinde yapılandır: - **Prettier**: Edit'ten sonra JS/TS dosyalarını otomatik formatla - **TypeScript check**: `.ts`/`.tsx` dosyalarını düzenledikten sonra `tsc` çalıştır ====================================================================== TITLE: TypeScript/JavaScript Testing TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/tr/rules/typescript/testing.md] > Bu dosya [common/testing.md](../common/testing.md) dosyasını TypeScript/JavaScript'e özgü içerikle genişletir. Kritik kullanıcı akışları için E2E test framework'ü olarak **Playwright** kullan. - **e2e-runner** - Playwright E2E testing uzmanı ====================================================================== TITLE: ECC v1.10.0 Release Notes TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/releases/1.10.0/release-notes.md] ECC v1.10.0 is a surface-sync and operator-lane release. The goal was to make the public repo, plugin metadata, install paths, and ecosystem story reflect the actual live state of the project again, while continuing to ship the operator workflows and media tooling that grew around the core harness layer. ====================================================================== TITLE: X Thread Draft — ECC v1.10.0 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/releases/1.10.0/x-thread.md] ECC crossed 140K stars and the public surface had drifted too far from the actual repo. so v1.10.0 is the sync release. 38 agents 156 skills 72 commands plugin metadata fixed install surfaces corrected docs and release story brought back in line with the live repo also shipped the operator / media lane that grew out of real usage: ====================================================================== TITLE: LinkedIn Draft - ECC v1.8.0 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/releases/1.8.0/linkedin-post.md] ECC v1.8.0 is now focused on harness performance at the system level. This release improves: - hook reliability and lifecycle behavior - eval-driven engineering workflows - operator tooling for autonomous loops - cross-platform support for Claude Code, Cursor, OpenCode, and Codex We also shipped NanoClaw v2 with stronger session operations for real workflow usage. ====================================================================== TITLE: ECC v1.8.0 Release Notes TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/releases/1.8.0/release-notes.md] ECC v1.8.0 positions the project as an agent harness performance system, not just a config bundle. - Stabilized hooks and lifecycle behavior. - Expanded eval and loop operations surface. - Upgraded NanoClaw for operational use. - Improved cross-harness parity (Claude Code, Cursor, OpenCode, Codex). 1. Validate hook profile defaults in your environment. ====================================================================== TITLE: X Quote Draft - Plankton / De-slop Workflow TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/releases/1.8.0/x-quote-plankton-deslop.md] The quality gate model matters. In v1.8.0 we pushed harder on write-time quality enforcement, deterministic checks, and cleaner loop recovery so agents converge faster with less noise. ====================================================================== TITLE: Seu Papel TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/agents/architect.md] Você é um arquiteto de software sênior especializado em design de sistemas escaláveis e manuteníveis. - Projetar arquitetura de sistemas para novas funcionalidades - Avaliar trade-offs técnicos - Recomendar padrões e boas práticas - Identificar gargalos de escalabilidade - Planejar para crescimento futuro ====================================================================== TITLE: Resolvedor de Erros de Build TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/agents/build-error-resolver.md] Você é um especialista em resolução de erros de build. Sua missão é fazer os builds passarem com o mínimo de alterações — sem refatorações, sem mudanças de arquitetura, sem melhorias. 1. **Resolução de Erros TypeScript** — Corrigir erros de tipo, problemas de inferência, restrições de generics 2. **Correção de Erros de Build** — Resolver falhas de compilação, resolução de módulos ====================================================================== TITLE: Revisor de Banco de Dados TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/agents/database-reviewer.md] Você é um especialista em PostgreSQL focado em otimização de queries, design de schema, segurança e performance. Sua missão é garantir que o código de banco de dados siga boas práticas, previna problemas de performance e mantenha integridade dos dados. Incorpora padrões das boas práticas postgres do Supabase (crédito: equipe Supabase). ====================================================================== TITLE: Executor de Testes E2E TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/agents/e2e-runner.md] Você é um especialista em testes end-to-end. Sua missão é garantir que jornadas críticas de usuário funcionem corretamente criando, mantendo e executando testes E2E abrangentes com gerenciamento adequado de artefatos e tratamento de testes instáveis. 1. **Criação de Jornadas de Teste** — Escrever testes para fluxos de usuário (preferir Agent Browser, fallback para Playwright) ====================================================================== TITLE: Comando E2E TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/commands/e2e.md] Este comando invoca o agente **e2e-runner** para gerar, manter e executar testes end-to-end usando Playwright. 1. **Gerar Jornadas de Teste** - Cria testes Playwright para fluxos de usuário 2. **Rodar Testes E2E** - Executa testes em múltiplos navegadores 3. **Capturar Artefatos** - Screenshots, vídeos, traces em falhas ====================================================================== TITLE: Comando Eval TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/commands/eval.md] Gerencie o fluxo de desenvolvimento orientado por evals. `/eval [define|check|report|list] [feature-name]` `/eval define feature-name` Crie uma nova definição de eval: 1. Crie `.claude/evals/feature-name.md` com o template: ```markdown Created: $(date) - [ ] [Descrição da capacidade 1] - [ ] [Descrição da capacidade 2] ====================================================================== TITLE: Build e Correção Go TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/commands/go-build.md] Este comando invoca o agente **go-build-resolver** para corrigir erros de build Go incrementalmente com mudanças mínimas. 1. **Rodar Diagnósticos**: Executa `go build`, `go vet`, `staticcheck` 2. **Parsear Erros**: Agrupa por arquivo e ordena por severidade 3. **Corrigir Incrementalmente**: Um erro por vez ====================================================================== TITLE: Comando TDD Go TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/commands/go-test.md] Este comando impõe a metodologia de desenvolvimento orientado a testes para código Go usando padrões idiomáticos de teste em Go. 1. **Definir Tipos/Interfaces**: Estrutura assinaturas de função primeiro 2. **Escrever Table-Driven Tests**: Criar casos de teste abrangentes (RED) 3. **Rodar Testes**: Verificar que os testes falham pelo motivo certo ====================================================================== TITLE: /learn - Extrair Padrões Reutilizáveis TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/commands/learn.md] Analise a sessão atual e extraia padrões que valem ser salvos como skills. Rode `/learn` em qualquer ponto da sessão quando você tiver resolvido um problema não trivial. Procure por: 1. **Padrões de Resolução de Erro** - Qual erro ocorreu? - Qual foi a causa raiz? - O que corrigiu? - Isso é reutilizável para erros semelhantes? ====================================================================== TITLE: Comando Plan TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/commands/plan.md] Este comando invoca o agente **planner** para criar um plano abrangente de implementação antes de escrever qualquer código. 1. **Reafirmar Requisitos** - Esclarecer o que precisa ser construído 2. **Identificar Riscos** - Levantar problemas e bloqueios potenciais 3. **Criar Plano em Etapas** - Quebrar implementação em fases ====================================================================== TITLE: Configuração de Package Manager TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/commands/setup-pm.md] Configure seu package manager preferido para este projeto ou globalmente. ```bash node scripts/setup-package-manager.js --detect node scripts/setup-package-manager.js --global pnpm node scripts/setup-package-manager.js --project bun node scripts/setup-package-manager.js --list ``` Ao determinar qual package manager usar, esta ordem é verificada: ====================================================================== TITLE: Comando TDD TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/commands/tdd.md] Este comando invoca o agente **tdd-guide** para impor a metodologia de desenvolvimento orientado a testes. 1. **Estruturar Interfaces** - Definir tipos/interfaces primeiro 2. **Gerar Testes Primeiro** - Escrever testes que falham (RED) 3. **Implementar Código Mínimo** - Escrever apenas o suficiente para passar (GREEN) ====================================================================== TITLE: Atualizar Documentação TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/commands/update-docs.md] Sincronize a documentação com o codebase, gerando a partir de arquivos fonte da verdade. | Source | Generates | |--------|-----------| | `package.json` scripts | Available commands reference | | `.env.example` | Environment variable documentation | | `openapi.yaml` / route files | API endpoint reference | ====================================================================== TITLE: Comando Verification TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/commands/verify.md] Rode verificação abrangente no estado atual do codebase. Execute a verificação nesta ordem exata: 1. **Build Check** - Rode o comando de build deste projeto - Se falhar, reporte erros e PARE 2. **Type Check** - Rode o TypeScript/type checker - Reporte todos os erros com file:line 3. **Lint Check** - Rode o linter ====================================================================== TITLE: Django REST API — CLAUDE.md de Projeto TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/examples/django-api-CLAUDE.md] > Exemplo real para uma API Django REST Framework com PostgreSQL e Celery. > Copie para a raiz do seu projeto e customize para seu serviço. **Stack:** Python 3.12+, Django 5.x, Django REST Framework, PostgreSQL, Celery + Redis, pytest, Docker Compose **Arquitetura:** Design orientado a domínio com apps por domínio de negócio. DRF para camada de API, Celery para tarefas assíncronas, pytest para tes ====================================================================== TITLE: Go Microservice — CLAUDE.md de Projeto TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/examples/go-microservice-CLAUDE.md] > Exemplo real para um microserviço Go com PostgreSQL, gRPC e Docker. > Copie para a raiz do seu projeto e customize para seu serviço. **Stack:** Go 1.22+, PostgreSQL, gRPC + REST (grpc-gateway), Docker, sqlc (SQL type-safe), Wire (injeção de dependência) **Arquitetura:** Clean architecture com camadas domain, repository, service e handler. gRPC como transporte principal com gateway REST para clie ====================================================================== TITLE: Aplicação SaaS — CLAUDE.md de Projeto TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/examples/saas-nextjs-CLAUDE.md] > Exemplo real para uma aplicação SaaS com Next.js + Supabase + Stripe. > Copie para a raiz do seu projeto e customize para sua stack. **Stack:** Next.js 15 (App Router), TypeScript, Supabase (auth + DB), Stripe (billing), Tailwind CSS, Playwright (E2E) **Arquitetura:** Server Components por padrão. Client Components apenas para interatividade. API routes para webhooks e server actions para mutaçõ ====================================================================== TITLE: Estilo de Código TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/rules/coding-style.md] SEMPRE crie novos objetos, NUNCA modifique os existentes: ``` // Pseudocódigo ERRADO: modificar(original, campo, valor) → altera o original in-place CORRETO: atualizar(original, campo, valor) → retorna nova cópia com a alteração ``` Justificativa: Dados imutáveis previnem efeitos colaterais ocultos, facilita a depuração e permite concorrência segura. ====================================================================== TITLE: Sistema de Hooks TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/rules/hooks.md] - **PreToolUse**: Antes da execução da ferramenta (validação, modificação de parâmetros) - **PostToolUse**: Após a execução da ferramenta (auto-formatação, verificações) - **Stop**: Quando a sessão termina (verificação final) Use com cautela: - Habilite para planos confiáveis e bem definidos - Desabilite para trabalho exploratório ====================================================================== TITLE: Padrões Comuns TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/rules/patterns.md] Ao implementar novas funcionalidades: 1. Buscar projetos skeleton bem testados 2. Usar agentes paralelos para avaliar opções: - Avaliação de segurança - Análise de extensibilidade - Pontuação de relevância - Planejamento de implementação 3. Clonar a melhor opção como fundação 4. Iterar dentro da estrutura comprovada ====================================================================== TITLE: Diretrizes de Segurança TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/pt-BR/rules/security.md] Antes de QUALQUER commit: - [ ] Sem segredos hardcoded (chaves de API, senhas, tokens) - [ ] Todas as entradas do usuário validadas - [ ] Prevenção de injeção SQL (queries parametrizadas) - [ ] Prevenção de XSS (HTML sanitizado) - [ ] Proteção CSRF habilitada - [ ] Autenticação/autorização verificada - [ ] Rate limiting em todos os endpoints ====================================================================== TITLE: 보안 리뷰어 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/ko-KR/agents/security-reviewer.md] 웹 애플리케이션의 취약점을 식별하고 수정하는 보안 전문 에이전트입니다. 보안 문제가 프로덕션에 도달하기 전에 방지하는 것이 목표입니다. 1. **취약점 감지** — OWASP Top 10 및 일반적인 보안 문제 식별 2. **시크릿 감지** — 하드코딩된 API 키, 비밀번호, 토큰 찾기 3. **입력 유효성 검사** — 모든 사용자 입력이 적절히 소독되는지 확인 4. **인증/인가** — 적절한 접근 제어 확인 5. **의존성 보안** — 취약한 npm 패키지 확인 6. **보안 모범 사례** — 안전한 코딩 패턴 강제 ```bash npm audit --audit-level=high ====================================================================== TITLE: Go Microservice — 프로젝트 CLAUDE.md TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/ko-KR/examples/go-microservice-CLAUDE.md] > PostgreSQL, gRPC, Docker를 사용하는 Go 마이크로서비스의 실전 예시입니다. > 프로젝트 루트에 복사하여 서비스에 맞게 커스터마이즈하세요. **기술 스택:** Go 1.22+, PostgreSQL, gRPC + REST (grpc-gateway), Docker, sqlc (타입 안전 SQL), Wire (의존성 주입) **아키텍처:** domain, repository, service, handler 레이어로 구성된 클린 아키텍처. gRPC를 기본 전송 프로토콜로 사용하고, 외부 클라이언트를 위한 REST gateway 제공. ====================================================================== TITLE: SaaS 애플리케이션 — 프로젝트 CLAUDE.md TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/ko-KR/examples/saas-nextjs-CLAUDE.md] > Next.js + Supabase + Stripe SaaS 애플리케이션을 위한 실제 사용 예제입니다. > 프로젝트 루트에 복사한 후 기술 스택에 맞게 커스터마이즈하세요. **기술 스택:** Next.js 15 (App Router), TypeScript, Supabase (인증 + DB), Stripe (결제), Tailwind CSS, Playwright (E2E) **아키텍처:** 기본적으로 Server Components 사용. Client Components는 상호작용이 필요한 경우에만 사용. API route는 webhook용, Server Action은 mutation용. ====================================================================== TITLE: 클라우드 및 인프라 보안 스킬 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/ko-KR/skills/security-review/cloud-infrastructure-security.md] | name | description | |------|-------------| | cloud-infrastructure-security | 클라우드 플랫폼 배포, 인프라 구성, IAM 정책 관리, 로깅/모니터링 설정, CI/CD 파이프라인 구현 시 이 스킬을 사용하세요. 모범 사례에 맞춘 클라우드 보안 체크리스트를 제공합니다. | 이 스킬은 클라우드 인프라, CI/CD 파이프라인, 배포 구성이 보안 모범 사례를 따르고 업계 표준을 준수하도록 보장합니다. - 클라우드 플랫폼(AWS, Vercel, Railway, Cloudflare)에 애플리케이션 배포 시 ====================================================================== TITLE: PM2 初期化 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/ja-JP/commands/pm2.md] プロジェクトを自動分析し、PM2サービスコマンドを生成します。 **コマンド**: `$ARGUMENTS` --- 1. PM2をチェック(欠落している場合は`npm install -g pm2`でインストール) 2. プロジェクトをスキャンしてサービスを識別(フロントエンド/バックエンド/データベース) 3. 設定ファイルと個別のコマンドファイルを生成 --- | タイプ | 検出 | デフォルトポート | |------|-----------|--------------| | Vite | vite.config.* | 5173 | | Next.js | next.config.* | 3000 | ====================================================================== TITLE: パッケージマネージャーの設定 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/ja-JP/commands/setup-pm.md] このプロジェクトまたはグローバルで優先するパッケージマネージャーを設定します。 ```bash node scripts/setup-package-manager.js --detect node scripts/setup-package-manager.js --global pnpm node scripts/setup-package-manager.js --project bun node scripts/setup-package-manager.js --list ``` 使用するパッケージマネージャーを決定する際、以下の順序でチェックされます: 1. **環境変数**: `CLAUDE_PACKAGE_MANAGER` ====================================================================== TITLE: プラグインとマーケットプレイス TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/ja-JP/plugins/README.md] プラグインは新しいツールと機能でClaude Codeを拡張します。このガイドではインストールのみをカバーしています - いつ、なぜ使用するかについては[完全な記事](https://x.com/affaanmustafa/status/2012378465664745795)を参照してください。 --- マーケットプレイスはインストール可能なプラグインのリポジトリです。 ```bash claude plugin marketplace add https://github.com/anthropics/claude-plugins-official claud plugin marketplace add https://github.com/mixedbread-ai/mgrep ====================================================================== TITLE: Django 検証ループ TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/ja-JP/skills/django-verification/SKILL.md] PR前、大きな変更後、デプロイ前に実行して、Djangoアプリケーションの品質とセキュリティを確保します。 ```bash python --version # プロジェクト要件と一致すること which python pip list --outdated python -c "import os; import environ; print('DJANGO_SECRET_KEY set' if os.environ.get('DJANGO_SECRET_KEY') else 'MISSING: DJANGO_SECRET_KEY')" ``` 環境が誤って構成されている場合は、停止して修正します。 ```bash ====================================================================== TITLE: フロントエンド開発パターン TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/ja-JP/skills/frontend-patterns/SKILL.md] React、Next.js、高性能ユーザーインターフェースのためのモダンなフロントエンドパターン。 ```typescript // PASS: GOOD: Component composition interface CardProps { children: React.ReactNode variant?: 'default' | 'outlined' } export function Card({ children, variant = 'default' }: CardProps) { return
{children}
====================================================================== TITLE: Javaコーディング標準 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/ja-JP/skills/java-coding-standards/SKILL.md] Spring Bootサービスにおける読みやすく保守可能なJava(17+)コードの標準。 - 巧妙さよりも明確さを優先 - デフォルトで不変; 共有可変状態を最小化 - 意味のある例外で早期失敗 - 一貫した命名とパッケージ構造 ```java // PASS: クラス/レコード: PascalCase public class MarketService {} public record Money(BigDecimal amount, Currency currency) {} // PASS: メソッド/フィールド: camelCase private final MarketRepository marketRepository; ====================================================================== TITLE: クラウドおよびインフラストラクチャセキュリティスキル TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/ja-JP/skills/security-review/cloud-infrastructure-security.md] | name | description | |------|-------------| | cloud-infrastructure-security | クラウドプラットフォームへのデプロイ、インフラストラクチャの設定、IAMポリシーの管理、ロギング/モニタリングの設定、CI/CDパイプラインの実装時にこのスキルを使用します。ベストプラクティスに沿ったクラウドセキュリティチェックリストを提供します。 | このスキルは、クラウドインフラストラクチャ、CI/CDパイプライン、デプロイメント設定がセキュリティのベストプラクティスに従い、業界標準に準拠することを保証します。 - クラウドプラットフォーム（AWS、Vercel、Railway、Cloudflare）へのアプリケーションのデプロイ ====================================================================== TITLE: Spring Boot 検証ループ TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/everything-claude-code/docs/ja-JP/skills/springboot-verification/SKILL.md] PR前、大きな変更後、デプロイ前に実行します。 ```bash mvn -T 4 clean verify -DskipTests ./gradlew clean assemble -x test ``` ビルドが失敗した場合は、停止して修正します。 Maven（一般的なプラグイン）: ```bash mvn -T 4 spotbugs:check pmd:check checkstyle:check ``` Gradle（設定されている場合）: ```bash ./gradlew checkstyleMain pmdMain spotbugsMain ``` ```bash mvn -T 4 test mvn jacoco:report # 80%以上のカバレッジを確認 ====================================================================== TITLE: Technical Specification TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/tools/ccdk/templates/docs/ai-context/spec.md] ====================================================================== TITLE: Contributing TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/infrastructure/mcp/tradingview-cdp/CONTRIBUTING.md] Thanks for your interest in contributing to tradingview-mcp. This tool is a **local bridge** between Claude Code and the TradingView Desktop app running on your machine. All contributions must stay within this scope. - Improving reliability of existing tools (better selectors, error handling, timeouts) ====================================================================== TITLE: TradingView MCP TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/infrastructure/mcp/tradingview-cdp/README.md] MCP server for TradingView Desktop — 78 tools to read, control, and automate charts via Chrome DevTools Protocol. Works with **Claude Code**, **Codex**, and **Claude Desktop**. **[Leer en espanol](README.es.md)** Built on top of [tradingview-mcp](https://github.com/tradesdontlie/tradingview-mcp) by [@tradesdontlie](https://github.com/tradesdontlie) and the [Jackson fork](https://github.com/LewisWJ ====================================================================== TITLE: Resumen Sesión de Trading — 05/04/2026 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/docs/contexto/resumen_sesion_05042026.md] --- - Death Cross activo - RSI 44 bajando - Precio debajo de EMA21 ($68,263), EMA50 ($70,364) y EMA200 ($84,223) - Targets bajistas: $66,063 → $65,236 → $62,660-$63,600 (OB LuxAlgo institucional) - Precio ~$67,091 comprimido entre EMA21 y FVG soporte - Bollinger Bands en squeeze — breakout inminente - MACD histograma bullish pero CVD bajando ====================================================================== TITLE: CONTEXTO COMPLETO — SESION 12/04/2026 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/docs/contexto/sesion_12042026_contexto_completo.md] --- - Balance: **100,777 VST** (inicio sesion 05/04: 99,971 VST → ganancia neta +806 VST) - Posiciones abiertas: 0 (todo cerrado) - Exchange: BingX demo VST - Simbolo operado: BTCUSDT perpetual - Modo cobertura: activo - Leverage default: 20x - Riesgo por operacion: 5,000 VST - SL: NO usa SL tradicional — usa liquidacion calculada sobre stop hunts ICT fadi ====================================================================== TITLE: feedback_trading TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/docs/memoria-claude-code/feedback_trading.md] **No usar Yahoo Finance para precios** — "usa TradingView no uses Yahoo, descarta Yahoo" **Why:** Yahoo Finance tiene datos limitados (60 dias intraday) y precios desfasados vs BingX. **How to apply:** Siempre usar mcp__tradingview-atila o mcp__tradingview-cdp para precios. **No todo a 20x** — pregunto "porque todos estan a 20x?" ====================================================================== TITLE: MEMORY TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/docs/memoria-claude-code/MEMORY.md] - [METODO TRADING COMPLETO](trading_method_complete.md) — LEER SIEMPRE antes de evaluar trades. BOAT, EMAs, Pirenest, OrderFlow, apalancamiento, checklist - [PLAN SISTEMA CONFLUENCIA](plan_sistema_confluencia.md) — PARCIALMENTE HECHO. Motor scoring 8 componentes funciona. Falta FreqTrade + parser mejorado. ====================================================================== TITLE: ESTADO: COMPLETADOS — 14/abril/2026 TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/docs/memoria-claude-code/plan_ajustes_pendientes.md] Todos los 8 ajustes fueron implementados. Este archivo es historico. 1. **Separar Discord y Telegram** — Ya estaba hecho (scripts independientes) 2. **bingx_sync_orders.py** — Sincroniza ordenes, limpia duplicados 3. **trailing_manager.py** — ATR + swing lows + muros orderbook, trail progresivo 4. **Alerta zonas Nousi** — En auto_analyzer.py, alerta si precio a <0.5% de zona ====================================================================== TITLE: ESTADO: PARCIALMENTE CONSTRUIDO (15/abril/2026) TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/docs/memoria-claude-code/plan_sistema_confluencia.md] - confluence_engine.py — motor de scoring 8 componentes (0-100) - senales_buffer.json — buffer de senales por analista - auto_analyzer.py — loop principal que monitorea y analiza - senales_historial.py — acumula senales, marca win/loss - coinglass_client.py — datos de derivados integrados - nousi_orderbook_validator.py — valida zonas con orderbook ====================================================================== TITLE: project_trading TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/docs/memoria-claude-code/project_trading.md] **Estado al 2026-04-15:** 19 posiciones abiertas en BingX DEMO (endpoint VST): - GOLD LONG 0.25 @ $4,769 | Lev 18x | PnL +$8.70 (+11.6%) — MEJOR posicion - SP500 SHORT 0.001 @ $6,920 | Lev 20x | PnL -$0.08 (-29.3%) — SL ajustado a $7,030 - 17 altcoins Wizards ALPHA (todas LONG, mixtas) - En profit: APT (+10.8%), LTC (+6.1%), XRP (+5%), BNB (+8.3%), HYPE (+1.4%), TIA (+7.4%), OP (+2.7%) ====================================================================== TITLE: reference_external TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/docs/memoria-claude-code/reference_external.md] **Discord — CRYPTOWZRD:** - wzrd-chat-crypto: 639277011164200960 - wzrd-chat-forex: 711688556669829130 - wzrd-crypto-setups: 773537812233715713 - wzrd-forex-setups: 780146729320185856 - User token para leer (REST API, poll cada 5s) **Telegram — Grupos VIP:** - BitLobo Satoshi VIP: -1002019332957 - BitLobo TRADING Proyecciones: -1001282984692 ====================================================================== TITLE: METODO DE APALANCAMIENTO Y MODO AISLADO (REGLA #1) TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/docs/memoria-claude-code/trading_method_complete.md] ``` PRECIO ACTUAL (entry) | | ← trade corriendo, trailing protege arriba | STOP HUNT / SWING LOW / MURO ← donde barrieron liquidez | | ← MARGEN/COLCHON (ATR segun volatilidad de la moneda) | PUNTO DE LIQUIDACION ← aca te saca si realmente rompio todo ``` **FORMULA:** ``` Leverage = 100 / (distancia_al_stop_hunt% + colchon_ATR%) ====================================================================== TITLE: Estrategia Final para Oro: v8 Stop Hunt (LA MEJOR) TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/docs/memoria-claude-code/trading_oro_research.md] **Pine Script:** `oro_v8_stophunt.pine` en F:\gaston-workspace\pine-scripts\ **Resultado en OANDA:XAUUSD 4H:** PF 1.70, WR 67%, DD 1.76% **Walk-forward:** WFER 156% (APROBADA) 1. ORB (Opening Range Breakout) 60 min 2. EMA200 + EMA50 filtro de tendencia 3. Awesome Oscillator confirmacion momentum 4. BB TP parcial 50% en Bollinger upper ====================================================================== TITLE: user_gaston TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/docs/memoria-claude-code/user_gaston.md] Gaston es un trader argentino que opera crypto (BTC, altcoins) y commodities (oro, petroleo) en BingX demo. **Metodo de trading:** - Entra en modo AISLADO con lo que puede perder (ej: $100 por trade) - El SL de la senal = punto de liquidacion (no SL separado) - Leverage = 100% / distancia_al_SL% (consecuencia del SL, no eleccion arbitraria) ====================================================================== TITLE: Notas de Mindpm - Exportadas TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/docs/memoria-permanente/mindpm_NOTAS.md] _Última actualización: 2026-03-29_ --- Sistema de inteligencia de mercado que aprende solo observando traders. NO es bot de señales, NO usa indicadores técnicos (no RSI, no Bollinger). Observa comportamiento colectivo de 50,000+ traders. - OpenClaw 24/7 en PC de oficina - HEARTBEAT cada 15min - CRON 8AM: análisis matutino (recolecta de 4 exchanges, filtros, scoring dinámico) ====================================================================== TITLE: Smart Follower System v2 - RESUMEN DEL PROYECTO TYPE: context | IMPORTANCE: 4 TAGS: ['proyecto:gaston-trading', 'doc-subproyecto'] [gaston-trading/docs/memoria-permanente/SMART_FOLLOWER_SYSTEM_RESUMEN.md] --- **No es un bot de señales. Es un sistema de inteligencia de mercado que aprende solo.** ``` Observa comportamiento colectivo de TODOS los traders de TODOS los exchanges ↓ Aprende patrones sistémicos del mercado ↓ Detecta correlaciones, ciclos, condiciones favorables ↓ Te sugiere operaciones con contexto completo ====================================================================== TITLE: polymarket-mcp:bug-leaderboard-endpoint-404 TYPE: bugfix | IMPORTANCE: 4 TAGS: ['polymarket', 'data-api', 'bug', 'endpoint', '404'] El endpoint /leaderboard del Data API (https://data-api.polymarket.com/leaderboard) retorna 404. La doc oficial menciona la ruta pero parece haberse migrado o requerir parametros distintos. Otros endpoints del Data API funcionan OK (positions, value, trades). Why: El MCP polymarket-weather tiene tool pmd_get_leaderboard que falla con HttpError 404. How to apply: Investigar la nueva ruta del leaderboard: probar /v2/leaderboard, /traders/leaderboard, o consultar docs.polymarket.com/api-reference. Ajustar core/polymarket/data.py:get_leaderboard. No bloqueante para Fase 1. ====================================================================== TITLE: Bienvenido TYPE: context | IMPORTANCE: 3 TAGS: ['vault:llm', 'nota-wiki'] [llm-wiki/Gastoon Workspace/Bienvenido.md] Esta es su nueva *bóveda*. Cree una nota de algo, [[cree un enlace]], o pruebe [el Importador](https://help.obsidian.md/Plugins/Importer)! Cuando esté listo, borre esta nota y aprópiese de la bóveda ====================================================================== TITLE: Sin título TYPE: context | IMPORTANCE: 3 TAGS: ['vault:llm', 'nota-wiki'] [llm-wiki/Gastoon Workspace/Sin título.md] ====================================================================== TITLE: CLAUDE.md - Reglas del Agente Wiki TYPE: context | IMPORTANCE: 3 TAGS: ['vault:trading', 'nota-wiki'] [trading-wiki/CLAUDE.md] Eres el mantenedor de este LLM Wiki. Sigue estas reglas. - index.md → Índice principal (SIEMPRE navegar desde acá) - aw/ → Fuentes crudas sin procesar - wiki/temas/ → Conceptos y temas - wiki/proyectos/ → Páginas por proyecto - wiki/entidades/ → Personas, organizaciones ====================================================================== TITLE: 📈 Gaston Trading - Índice TYPE: context | IMPORTANCE: 3 TAGS: ['vault:trading', 'nota-wiki'] [trading-wiki/index.md] - [[bingx-trading]] — Trading demo BingX (leverage, órdenes, riesgo) - [[tradingview-cdp]] — Control TradingView Desktop vía CDP - [[tradingview-atila]] — Análisis técnico vía TradingView Atila MCP - [[trade-analyze]] — Análisis de trading estructurado (checklist + técnico + resumen) ====================================================================== TITLE: Ingest — 2026-04-22 TYPE: context | IMPORTANCE: 3 TAGS: ['vault:trading', 'nota-wiki'] [trading-wiki/log/2026-04-22-ingest.md] - Leídos: `plugin.json`, `.mcp.json`, 5 SKILL.md - Creadas: - `wiki/proyectos/codex-trading-plugin.md` - `wiki/temas/bingx-trading.md` - `wiki/temas/tradingview-cdp.md` - `wiki/temas/tradingview-atila.md` - `wiki/temas/trade-analyze.md` - `wiki/temas/trading-session.md` ====================================================================== TITLE: Codex Trading Plugin TYPE: context | IMPORTANCE: 3 TAGS: ['vault:trading', 'nota-wiki'] [trading-wiki/wiki/proyectos/codex-trading-plugin.md] > Fuente: `F:\gaston-workspace\codex-trading-plugin` Plugin de Codex (OpenAI) que empaqueta el workflow de trading de Gastón como skills + MCP servers. Incluye 5 skills especializadas y 5 servidores MCP para integrar TradingView, BingX, análisis de mercado y productividad. ====================================================================== TITLE: Gaston Trading TYPE: context | IMPORTANCE: 3 TAGS: ['vault:trading', 'nota-wiki'] [trading-wiki/wiki/proyectos/gaston-trading.md] > Fuente: `F:\gaston-workspace\gaston-trading` Sistema multi-activo de trading automatizado con lectura de señales, análisis automático y backtesting. Diseñado para correr en Windows con soporte para IA local. Cubre crypto (BTC, ETH, SOL, alts) y commodities (oro, petróleo). ====================================================================== TITLE: TradingView Atila TYPE: context | IMPORTANCE: 3 TAGS: ['vault:trading', 'nota-wiki'] [trading-wiki/wiki/temas/tradingview-atila.md] > Fuente: `F:\gaston-workspace\codex-trading-plugin\skills\tradingview-atila\SKILL.md` Skill que utiliza el MCP de TradingView Atila para obtener contexto de mercado rápido, análisis técnico, precios y noticias financieras. 51 herramientas disponibles para contexto multi-timeframe. ====================================================================== TITLE: TradingView CDP TYPE: context | IMPORTANCE: 3 TAGS: ['vault:trading', 'nota-wiki'] [trading-wiki/wiki/temas/tradingview-cdp.md] > Fuente: `F:\gaston-workspace\codex-trading-plugin\skills\tradingview-cdp\SKILL.md` Skill para controlar TradingView Desktop localmente mediante el protocolo Chrome DevTools (CDP). Permite editar Pine Scripts, inspeccionar charts, ejecutar Strategy Tester y gestionar layouts multi-tab.