name: ops-deploy\ndescription: Deploy status across all projects. Shows ECS service versions, Vercel deployments, recent deploys, pending deploys, and CI/CD pipeline state.\nargument-hint: \"[project-alias|ecs|vercel|all]\"\nallowed-tools:
\nBefore executing, load available context:
\nPreferences: Read ${CLAUDE_PLUGIN_DATA_DIR:-$HOME/.claude/plugins/data/ops-ops-marketplace}/preferences.json
timezone — display all deploy timestamps in the correct timezoneDaemon health: Read ${CLAUDE_PLUGIN_DATA_DIR}/daemon-health.json
infra-monitor status — if not running, note that ECS data may be staleSecrets: AWS and Vercel credentials required.
\n$AWS_PROFILE / $AWS_ACCESS_KEY_ID → doppler secrets get AWS_ACCESS_KEY_ID --plain → vault query cmd from prefs$VERCEL_TOKEN → doppler secrets get VERCEL_TOKEN --plain → vault| Command | \nUsage | \nOutput | \n
|---|---|---|
aws ecs list-clusters --output json | \nAll ECS clusters | \n{clusterArns: [...]} | \n
aws ecs list-services --cluster <name> --output json | \nServices in cluster | \n{serviceArns: [...]} | \n
aws ecs describe-services --cluster <name> --services <arn> --output json | \nService health | \n{services: [{serviceName, status, runningCount, desiredCount, pendingCount}]} | \n
aws logs tail /ecs/<service> --since 1h --format short | \nECS logs | \nLog lines | \n
| Command | \nUsage | \nOutput | \n
|---|---|---|
gh run list --repo <owner/repo> --limit 5 --json status,conclusion,name,headBranch,createdAt,databaseId | \nCI runs | \nJSON array | \n
gh run view <id> --repo <repo> --log-failed | \nFailed CI logs | \nLog output | \n
gh run watch <run-id> --repo <repo> | \nStream CI run | \nLive output (use with Monitor) | \n
If CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1 is set, use Agent Teams when checking deploy platforms in parallel. This enables:
Team setup (only when flag is enabled):
\nTeamCreate(\"deploy-team\")\nAgent(team_name=\"deploy-team\", name=\"ecs-checker\", prompt=\"List all ECS clusters and describe service health, running/desired counts\")\nAgent(team_name=\"deploy-team\", name=\"vercel-checker\", prompt=\"List Vercel projects and recent deployments with status\")\nAgent(team_name=\"deploy-team\", name=\"ci-checker\", prompt=\"Check GitHub Actions runs across all registered repos for failures\")\nIf the flag is NOT set, use standard fire-and-forget subagents.
\n${CLAUDE_PLUGIN_ROOT}/bin/ops-infra 2>/dev/null || \\\naws ecs list-clusters --output json 2>/dev/null\nfor cluster in $(aws ecs list-clusters --output json 2>/dev/null | jq -r '.clusterArns[]'); do\n cluster_name=$(basename \"$cluster\")\n aws ecs list-services --cluster \"$cluster_name\" --output json 2>/dev/null | \\\n jq -r '.serviceArns[]' | while read svc; do\n aws ecs describe-services --cluster \"$cluster_name\" --services \"$svc\" \\\n --output json 2>/dev/null | jq '.services[] | {name: .serviceName, desired: .desiredCount, running: .runningCount, pending: .pendingCount, image: (.taskDefinition // \"unknown\"), status: .status}'\n done\ndone\nREGISTRY=\"${CLAUDE_PLUGIN_ROOT}/scripts/registry.json\"\n[ -f \"$REGISTRY\" ] || REGISTRY=\"${CLAUDE_PLUGIN_ROOT}/scripts/registry.example.json\"\nfor repo in $(jq -r '.projects[] | select(.gsd == true) | .repos[]' \"$REGISTRY\" 2>/dev/null); do\n echo \"=== $repo ===\"\n gh run list --repo \"$repo\" --limit 5 --json status,conclusion,name,headBranch,createdAt,databaseId 2>/dev/null\ndone\nUse mcp__claude_ai_Vercel__list_projects then mcp__claude_ai_Vercel__list_deployments for each project (limit 5 per project).
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n OPS ► DEPLOY STATUS — [timestamp]\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n\nECS SERVICES\n CLUSTER SERVICE D/R/P STATUS LAST DEPLOY\n ─────────────────────────────────────────────────────\n [cluster] [service] [x/x/x] ACTIVE [time ago]\n ...\n\nVERCEL DEPLOYMENTS\n PROJECT ENV STATUS COMMIT DEPLOYED\n ─────────────────────────────────────────────────────\n [project] production READY [sha] [time ago]\n ...\n\nCI/CD PIPELINE\n REPO BRANCH WORKFLOW STATUS AGE\n ─────────────────────────────────────────────────────\n example-api main Deploy API ✓ success 2h\n example-web dev Build ✗ failure 1h\n ...\n\nPENDING DEPLOYS (branch ready, not yet deployed)\n [repo] [branch] [PR#] [CI status] → needs merge to trigger\n\n──────────────────────────────────────────────────────\nAfter rendering, use batched AskUserQuestion calls (max 4 options each). Only show actions relevant to the current state (e.g., skip \"View logs for failing service\" if nothing is failing). If <=4 relevant actions, use a single call. If >4, batch:
\nAskUserQuestion call 1:
\n [View logs for [failing service]]\n [Trigger manual deploy for [project]]\n [View build logs for [failing CI run]]\n [More actions...]\nAskUserQuestion call 2 (only if \"More actions...\"):
\n [Check Vercel [project] runtime logs]\n [Open GitHub Actions for [repo]]\n [Back to dashboard]\nIf $ARGUMENTS has a project alias, show only that project's deploy info + last 10 CI runs + option to view logs.
For failing deploys: offer to view logs via mcp__claude_ai_Vercel__get_deployment_build_logs or ECS CloudWatch logs.
If user selects manual deploy (option b), confirm with AskUserQuestion before triggering:
Trigger deploy for [project]:\n Environment: [production/staging]\n Branch: [branch]\n Last commit: [sha] — [message]\n\n [Deploy now] [View diff since last deploy first] [Cancel]\nIf user selects to view logs, show the logs and use AskUserQuestion:
[Dispatch fix agent for this failure] [Redeploy] [Back to dashboard]\nWhen watching a deploy in progress, use Monitor to stream logs:
Monitor(command: \"gh run watch <run-id> --repo <repo>\")\nFor ECS deploys: Monitor(command: \"aws ecs wait services-stable --cluster <cluster> --services <service>\")
Use TaskCreate per project being deployed. Update with TaskUpdate as deploys succeed/fail.
When Vercel MCP tools are unavailable, use WebFetch with the Vercel API directly:
WebFetch(url: \"https://api.vercel.com/v6/deployments?projectId=<id>&limit=5\", headers: {\"Authorization\": \"Bearer $VERCEL_TOKEN\"})\n