davidcastagnetoa/skills

request_timeout_management

Timeout global por tipo de endpoint con cancelación limpia

First seen Mar 3, 2026

Installation

$ npx skills add davidcastagnetoa/skills --skill request_timeout_management

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

Also in this package

Other skills from davidcastagnetoa/skills · top by installs.

npx skills add davidcastagnetoa/skills

Browse all from davidcastagnetoa/skills

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Repository health

Stars 1
Default branch main
Open issues 0
Status Active

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 1,559 B
  • docs SUMMARY.md 93 B

History

  1. First seen on skills.sh
  2. First recorded snapshot · 9 installs

SKILL.md

requesttimeoutmanagement

Gestión de timeouts configurables por tipo de endpoint para garantizar que ninguna petición bloquee recursos indefinidamente. Alineado con el SLO de 8 segundos del pipeline KYC.

When to use

Usar en el apigatewayagent para establecer timeouts diferenciados por endpoint. El timeout total de verificación debe ser < 8s; endpoints de consulta < 2s.

Instructions

  1. Configurar timeouts en Nginx por location:

``nginx location /v1/verify { proxyreadtimeout 8s; proxyconnecttimeout 2s; proxysendtimeout 2s; } location /v1/status { proxyreadtimeout 2s; } ``

  1. Configurar timeout global de keepalive: keepalive_timeout 65s.
  2. En FastAPI, usar asyncio timeout para cancelación limpia:

``python async with asyncio.timeout(7.5): # 0.5s margen para response result = await orchestrator.verify(session) ``

  1. Si timeout, devolver 504 Gateway Timeout con session_id para retry.
  2. Registrar timeouts en métricas: gatewaytimeoutstotal{endpoint}.
  3. Alertar si tasa de timeouts > 1% en ventana de 5 minutos.

Notes

  • El timeout del gateway debe ser mayor que el del backend para evitar cortar respuestas válidas.
  • Implementar client-side timeout también en los SDKs de clientes.
  • Los timeouts deben ser configurables por entorno (dev: más largos, prod: estrictos).