apexscaleai Infra Optimization
Use this skill to improve deployment and runtime infrastructure with a bias toward measurable outcomes: faster builds, cheaper runtime, safer deploys, and better observability.
Do not assume an infra stack. Discover what's actually used in the repo first.
Quick Start
- Run discovery to map infra surface area:
scripts/discover_infra.py. - Fill the checklist using repo evidence (PASS/FAIL/UNKNOWN):
references/infra-checklist.md. - Propose 3-5 high ROI changes with expected impact and verification steps.
Workflow (Prescriptive)
- Discover the infra surface area (evidence-first).
- Look for infra code:
infrastructure/,infra/,terraform/,pulumi/,k8s/,helm/,docker-compose.yml,Dockerfile*. - Identify hosting targets:
- Web: Vercel, Netlify, AWS, GCP, self-hosted, containers, serverless, edge.
- Mobile: build pipelines (Fastlane/EAS), OTA updates, artifact signing, store release flow.
- Identify CI: GitHub Actions, CircleCI, Buildkite, etc.
- Identify observability: Sentry, Datadog, OpenTelemetry, LogRocket, custom logging.
- Establish current bottlenecks and budgets.
- Build time (CI minutes) and cache hit rate (if available).
- Bundle size and build output size (web + mobile).
- Runtime cost drivers (serverless invocations, DB queries, bandwidth/CDN).
- Reliability risks (no canary, no rollback plan, missing alerts).
- Apply the checklist.
- Use
references/infra-checklist.mdand mark each item PASS/FAIL/UNKNOWN with links to files.
- Recommend improvements (ranked).
- Provide a short list ranked by ROI:
- Expected impact (time/cost/risk)
- Complexity (low/med/high)
- Blast radius
- Verification steps
- If asked to implement, do it safely.
- Prefer incremental changes (one deploy pipeline/cache change at a time).
- Include a rollback path for infra changes.
Resources
- Checklist:
references/infra-checklist.md - Report template:
assets/infra-optimization-report.md - Discovery script:
scripts/discover_infra.py