estatewise-ops

# EstateWise Ops Invoke this skill manually for infrastructure, deployment-control, or operational workflow changes. ## Scope - `deployment-control/` - `docker/` - `kubernetes/` - `helm/` - `terraform/` - `aws/`, `azure/`, `gcp/`, `oracle-cloud/`, `hashicorp/` - `jenkins/`, `gitlab/` - root docs: `DEPLOYMENTS.md`, `DEVOPS.md` ## Guardrails - Do not run destructive deployment or cluster commands unless the task explicitly requires it. - Preserve rollout semantics for blue-green, canary, and rolling paths. - Keep environment-specific defaults backward compatible unless explicitly asked to change them. - Treat `deployment-control` as trusted-environment software unless the task adds auth/RBAC deliberately. ## Important EstateWise Details - `deployment-control/src/server.ts` builds shell commands for rollout and scale actions. - Job lifecycle semantics matter: `queued`, `running`, `succeeded`, `failed`. - UI is served separately via Nuxt on port `3000`; API defaults to `4100`. - MCP and agentic-ai deployments are often sidecar or bundled runtime patterns, not generic stateless web apps. - MCP is stdio-oriented; deployment docs must preserve that mental model. ## Validation ### Deployment Control ```bash cd deployment-control npm run build:api npm run build:ui ``` ### Docker / Compose docs or manifests - Validate referenced image names, env vars, and sidecar assumptions against existing files. ### Kubernetes / Helm / Terraform - Prefer static validation and focused manifest inspection unless the task explicitly requests live cluster actions. - Confirm values stay aligned with service ports, env vars, and container entrypoints used elsewhere in the repo. ## Documentation To Update - `deployment-control/README.md` - `docker/README.md` - `kubernetes/README.md` - `helm/estatewise/README.md` - `DEPLOYMENTS.md` - `DEVOPS.md` - cloud-specific READMEs when their deployment path changes ## Review Questions - Does the change increase operational blast radius? - Does it change the contract between deployment-control API and UI? - Does it change how MCP or agentic-ai is launched or wired? - Are auth/security assumptions still explicitly documented? - Are rollback and status-observability paths still intact?