メニュー
Deploy ssmd components to GKE prod. Handles git tagging, GitHub Actions image builds, manifest updates, and Flux reconciliation. Use when deploying connector, archiver, parquet-gen, DQ, data-ts, CLI, operator, worker, or other ssmd images after code changes.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
| name | ssmd-deploy |
| description | Deploy ssmd components to GKE prod. Handles git tagging, GitHub Actions image builds, manifest updates, and Flux reconciliation. Use when deploying connector, archiver, parquet-gen, DQ, data-ts, CLI, operator, worker, or other ssmd images after code changes. |
Deploy ssmd components from code change through to running in GKE prod.
ssmd images are built by GitHub Actions on tag push. Deployments are GitOps via Flux: update image tags in clusters/gke-prod/ manifests, push, and Flux reconciles.
There are two repos involved:
899bushwick/ssmd/) — source code, where tags trigger image builds| Component | Tag Format | Image | Manifest Path(s) |
|---|---|---|---|
| connector | v* (e.g. v0.9.11) | ghcr.io/aaronwald/ssmd-connector | clusters/gke-prod/apps/ssmd/generated/connector-{crypto,kraken-futures,polymarket}.yaml |
| archiver | v* (same tag as connector) | ghcr.io/aaronwald/ssmd-archiver | clusters/gke-prod/apps/ssmd/generated/archiver-{kalshi-crypto,kraken-futures,polymarket}.yaml |
| parquet-gen | parquet-gen-v* | ghcr.io/aaronwald/ssmd-parquet-gen | clusters/gke-prod/apps/ssmd/archiver/cronjob-parquet-gen.yaml |
| operator | operator-v* | ghcr.io/aaronwald/ssmd-operator | clusters/gke-prod/apps/ssmd/operator/deployment.yaml |
| data-ts | data-ts-v* | ghcr.io/aaronwald/ssmd-data-ts | clusters/gke-prod/apps/ssmd/data-ts/deployment.yaml |
| cli-ts | cli-ts-v* | ghcr.io/aaronwald/ssmd-cli-ts | (used in CronJobs: secmaster-crypto, kraken-sync, polymarket-sync, fees-daily, health-daily) |
| worker | worker-v* | ghcr.io/aaronwald/ssmd-worker | clusters/gke-prod/apps/ssmd/worker/deployment.yaml |
| signal-runner | signal-runner-v* | ghcr.io/aaronwald/ssmd-signal-runner | clusters/gke-prod/apps/ssmd/signals/ |
| Component | Tag Format | Image | Manifest Path |
|---|---|---|---|
| dq | dq-v* | ghcr.io/aaronwald/ssmd-dq | clusters/gke-prod/apps/ssmd/cronjobs/dq-daily.yaml |
Identify which components changed. Common groupings:
v*), parquet-gen (parquet-gen-v*)v*)dq-v*)parquet-gen-v*)operator-v*), also copy CRDs to clusters/gke-prod/apps/ssmd/operator/crds.yamlCheck the latest tags:
# For ssmd repo components
git -C <ssmd-path> tag --sort=-creatordate | grep "^v[0-9]" | head -3 # connector/archiver
git -C <ssmd-path> tag --sort=-creatordate | grep "^parquet-gen-v" | head -3
git -C <ssmd-path> tag --sort=-creatordate | grep "^operator-v" | head -3
git -C <ssmd-path> tag --sort=-creatordate | grep "^data-ts-v" | head -3
# For 899bushwick components
git -C <899bushwick-path> tag --sort=-creatordate | grep "^dq-v" | head -3
Increment the patch version (e.g. v0.9.10 -> v0.9.11).
# ssmd repo
git -C <ssmd-path> add <changed-files>
git -C <ssmd-path> commit -m "<type>: <description>"
git -C <ssmd-path> push origin main
# 899bushwick repo (for DQ/data changes)
git -C <899bushwick-path> add <changed-files>
git -C <899bushwick-path> commit -m "<type>: <description>"
git -C <899bushwick-path> push origin main
# ssmd repo tags
git -C <ssmd-path> tag <tag>
git -C <ssmd-path> push origin <tag>
# 899bushwick tags
git -C <899bushwick-path> tag <tag>
git -C <899bushwick-path> push origin <tag>
Multiple tags can be pushed at once: git push origin v0.9.11 parquet-gen-v0.7.0
gh run list --repo aaronwald/ssmd --limit 5 --json name,status,conclusion,headBranch
gh run list --repo aaronwald/899bushwick --limit 3 --json name,status,conclusion,headBranch
Poll until all relevant builds show "conclusion":"success". Rust builds take 3-5 minutes; Python/Deno builds take 1-2 minutes.
Update image tags in the manifest files listed in the Component Registry table above.
The image: field format is: ghcr.io/aaronwald/<image-name>:<version-without-prefix>
v0.9.11 -> image tag 0.9.11parquet-gen-v0.7.0 -> image tag 0.7.0dq-v0.8.0 -> image tag 0.8.0Important: Connector and archiver share the same version number (same v* tag triggers both builds).
git -C <899bushwick-path> add <manifest-files> ssmd # include ssmd submodule if updated
git -C <899bushwick-path> commit -m "deploy: <component> <version>[, <component> <version>...]"
git -C <899bushwick-path> push origin main
Flux auto-reconciles on git push. To force or check:
flux reconcile kustomization ssmd --with-source
kubectl get pods -n ssmd -w # watch for rolling restarts
v* tag). Even if only one changed, both get rebuilt.kubectl apply directly — all changes go through git + Flux.Components: connector v0.9.11, parquet-gen v0.7.0
Tags: v0.9.11, parquet-gen-v0.7.0 (in ssmd repo)
Manifests: 3 connector yamls, 3 archiver yamls, 1 parquet-gen cronjob
Components: dq v0.8.0
Tags: dq-v0.8.0 (in 899bushwick repo)
Manifests: 1 dq-daily cronjob
Components: operator v0.5.6
Tags: operator-v0.5.6 (in ssmd repo)
Extra step: copy CRDs from ssmd-operators to clusters/gke-prod/apps/ssmd/operator/crds.yaml
Manifests: 1 operator deployment + 1 crds.yaml
Orchestrate multi-stage pipelines for ssmd workflows. Pre-defined pipelines chain skills and agents in sequence — deploy, feature, hotfix, and dq-investigate. Use when user says "ssmdorch", "orchestrate", "pipeline", or named pipelines like "ssmdorch deploy", "ssmdorch feature", "ssmdorch hotfix", "ssmdorch dq".
Multi-agent orchestration for ssmd market data system tasks. Extends waldstorm with domain-specific experts for secmaster, data feeds, trading APIs, and data quality. Use when working on connectors, exchanges, NATS pipelines, market metadata, or user says "ssmdstorm", "market data task", "exchange work".
Multi-agent orchestration that analyzes tasks through specialized expert panels (security, performance, architecture, etc.), synthesizes prioritized actions, then plans and executes. Use when facing complex tasks requiring multiple perspectives, architectural decisions, new features needing security/performance/quality review, or user says "waldstorm", "expert review", "analyze this task".
This skill MUST be used before writing any implementation code — feature work, bug fixes, pipeline stages, data processing, API handlers, K8s manifests, or integration code. Enforces fail-loud patterns, input/output validation, connection verification, and pre-commit gates. Triggered automatically on any code writing task. Also use when user says "defensive", "fail-fast", "validate", "check failures", "harden".
Catalog of all ssmd health metrics — sources, query methods, labels, healthy/unhealthy thresholds. Covers data pipeline Prometheus metrics, HTTP endpoints, infrastructure components, DQ scores, and composite health scoring. Use when checking environment health, interpreting metrics, setting alert thresholds, or building health dashboards.
Procedures for running ssmd health checks and interpreting results. Covers quick triage, data pipeline health, infrastructure checks (NATS, Postgres, Redis), connector/archiver deep dives via port-forward, and Cloud Monitoring queries. Use when checking environment health, investigating degradation, or verifying health after deployments.