mlops-model-serving

📁 kentoshimizu/sw-agent-skills 📅 1 day ago
0
总安装量
1
周安装量
安装命令
npx skills add https://github.com/kentoshimizu/sw-agent-skills --skill mlops-model-serving

Agent 安装分布

amp 1
cline 1
opencode 1
cursor 1
continue 1
kimi-cli 1

Skill 文档

Mlops Model Serving

Overview

Use this skill to deploy models with predictable latency/error behavior and controlled rollout risk.

Scope Boundaries

  • Use this skill when the task matches the trigger condition described in description.
  • Do not use this skill when the primary task falls outside this skill’s domain.

Shared References

  • Serving SLO and rollout rules:
    • references/serving-slo-and-rollout-rules.md

Templates And Assets

  • Serving readiness checklist:
    • assets/serving-readiness-checklist.md

Inputs To Gather

  • Serving topology and traffic profile.
  • Latency/error SLO and error budget constraints.
  • Rollout strategy and rollback capability.
  • Observability and incident response expectations.

Deliverables

  • Serving architecture and rollout plan.
  • SLO-aligned guardrails and alert thresholds.
  • Readiness evidence and rollback criteria.

Workflow

  1. Define serving constraints and topology.
  2. Apply references/serving-slo-and-rollout-rules.md for rollout policy.
  3. Validate readiness with assets/serving-readiness-checklist.md.
  4. Execute staged rollout and monitor guardrails.
  5. Publish serving decision and residual risk ownership.

Quality Standard

  • Serving SLOs are measurable and enforced.
  • Rollout blast radius is controlled.
  • Rollback decisions are objective and fast.

Failure Conditions

  • Stop when serving cannot meet latency/reliability targets.
  • Stop when rollback path is unverified.
  • Escalate when rollout risk exceeds policy thresholds.