foundry-ops

Plugin foundry-ops from marketplace foundry in fedcal/foundry. Commands read from the repo's manifest — 1 month ago.

Skills (5)

• containerise

  Produce a hardened multi-stage Dockerfile plus a docker compose file for local development, matched to the stack detected in the repository (Node, Angular static, Java/Spring, Python). Use when a project needs its first container, when an image is oversized or slow to build, when it runs as root, or

• deploy-strategy

  Choose and configure the deployment strategy for a service (rolling update, blue-green or canary), size its parameters from measurements, and write the rollback procedure as a Foundry runbook. Use before the first production deploy, after a bad rollout, when adopting progressive delivery, or when no

• provision-environment

  Create or extend Terraform/OpenTofu for a new environment, with remote state and locking, secrets handling and least-privilege identity for CI and workloads. Use when adding staging or production, splitting a shared account, moving from local state to remote, or replacing long-lived cloud keys with

• release

  Run a release end to end - derive the version from the commit range and an API diff, generate the changelog, tag, publish and promote artefacts by digest, write release notes with the rollback classification, and verify after release. Use when cutting a release, promoting a build between environment

• scaffold-pipeline

  Generate a complete, hardened GitHub Actions pipeline for the project type detected in the repository (Node/Angular, Java with Maven or Gradle, containerised service) with build, test, quality gate, security scan, SBOM, release and environment-gated deploy. Use when a repository has no CI, when CI c

Subagents (6)

• cloud-architect

  Chooses the deployment target and its boundaries across AWS, Azure, GCP and PaaS (Vercel, Netlify, Fly.io, Render, Railway). Use before building infrastructure, when a team assumes Kubernetes by default, when comparing managed services, or when someone needs the cost shape and the exit cost of a cho

• container-engineer

  Writes and hardens container images. Use when a Dockerfile is missing, when images are large or slow to build, when a scanner reports vulnerabilities and nobody knows which ones matter, when a container runs as root, or when a service ignores SIGTERM and drops connections on redeploy.

• iac-engineer

  Terraform and OpenTofu at scale. Use when infrastructure code needs module boundaries, remote state and locking, environment separation without copy-paste, drift detection, plan review discipline, policy-as-code, or when deciding which apply operations are safe to automate and which must never be.

• kubernetes-engineer

  Designs Kubernetes workloads and rollout strategy. Use when sizing requests and limits, when pods are OOMKilled or CPU-throttled, when probes cause restart storms or 502s during deploys, when node drains hang, when choosing between rolling update, blue-green and canary, or when a rollback needs to b

• pipeline-engineer

  Designs, hardens and speeds up GitHub Actions pipelines. Use when a repository has no CI, when PR feedback is slow, when the workflow uses long-lived cloud secrets or unpinned third-party actions, or when build artefacts and SBOMs need to be published and attested.

• release-engineer

  Runs release trains. Use for version decisions under SemVer, conventional-commit changelogs, tagging, promoting a single artefact across environments by digest, using feature flags to decouple deploy from release, and coordinating a multi-service release with version-skew handling.

Similar plugins