ship
Deploy Trigger.dev tasks (trigger deploy) to staging, prod or a preview branch with a preflight, the right CI or integration pipeline, an explicit --env, and a verification step. Use when asked to deploy, ship or release the Trigger.dev tasks or jobs. Not for deploying the app itself to Vercel, Fly or another host.
- License 许可
- MIT
- 最近更新
- 2026-11-02
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- ship
- description
- Deploy Trigger.dev tasks (trigger deploy) to staging, prod or a preview branch with a preflight, the right CI or integration pipeline, an explicit --env, and a verification step. Use when asked to deploy, ship or release the Trigger.dev tasks or jobs. Not for deploying the app itself to Vercel, Fly or another host.
- allowed-tools
- Bash(trigger-ctx *) Bash(node *trigger-ctx.mjs *) Bash(node *trigger-ctx.mjs" *) Bash(git status *) Bash(git branch --show-current)
- argument-hint
- [staging|prod|preview]
03 SKILL.md 正文
Ship Trigger.dev tasks ($ARGUMENTS)
Project facts
!node "${CLAUDE_PLUGIN_ROOT}/scripts/trigger-ctx.mjs" detect --brief
Read ${CLAUDE_PLUGIN_ROOT}/skills/build/scenarios/deploy-ci.md first. Deploys touch production: the target environment is always explicit (--env), and a production deploy needs the user's yes in this conversation. If no environment was given, ask: staging, prod or preview.
Steps
- Preflight. Run
trigger-ctx scan --focus deploy --json. Fix or report: version drift betweentrigger.devand@trigger.dev/*(CI fails on a mismatch); imports whose extension is missing (browser, Prismamodeand a generate step, ffmpeg withfluent-ffmpegexternal, Python scripts, native modules asbuild.external);process.env.Xread underdirsthat is not set in the target environment (npx trigger.dev env list --env <e>, without--show-values); secrets in code or payloads; a defaultmaxDuration; the--env-filemisbelief (it only hydrates the CLI process). - Pipeline. Infer from
.github/workflowsandvercel.json; ask once if unclear: manual CLI · GitHub Actions (aTRIGGER_ACCESS_TOKENper environment, deploy through the devDependency script,--external-idwith the commit sha and the PR head sha on pull requests, a preview workflow listening foropened, synchronize, reopened, closed) · GitHub integration (config path when there is more than one config, a pre-build step for Prisma generate) · Vercel integration (needs the GitHub integration; removesyncVercelEnvVars) · preview branches (the integration or Actions, never both). Integrations and key presets are Cloud features: skip them on a self-hosted target. - Version skew. If the app is deployed separately from the tasks, set
TRIGGER_EXTERNAL_DEPLOYMENT_IDorconfigure({ externalDeploymentId }). Watch forpaths:filters, two apps sharing a project, and a staleTRIGGER_VERSION. - Deploy with the project's own CLI and an explicit environment:
pnpm exec trigger deploy --env <e>(or--prod; ornpx trigger deploy --env <e>). With aTRIGGER_ACCESS_TOKENthat is an environment key, an omitted--envtargets that key's environment, so still write it. Run--dry-runfirst when extensions changed. Onresource_exhaustedretry with--native-build. - Verify. MCP
list_deploys(explicit environment) shows the new version;get_current_workerin that environment lists the tasks; optionally a smoketrigger_taskin staging (ask first);get_report(health) for that environment. Show the version and the dashboard link. If anything could not be checked, say UNVERIFIED and why. - Runtime updates need a redeploy. A deployed task keeps the Node runtime and CA trust store it was built with. After a Trigger.dev security release (for example 4.7.7: Node security patches; 4.7.6: removed CA roots), redeploying on the newest CLI is what picks them up, except on a custom
image.base/image.buildBase(self-hostDEPLOY_BUILD_BASE_IMAGES): from 4.8.1 such a stage keeps the image's own trust store, so patching the image is the owner's job (self-hosting/overview). Mention it when the last deploy predates the release in question. - Rollback is
promote <previous version> --env <e>. If the project has chat agents, warn that the chat transcript format is roll-forward only.
Human gates (name them, don't work around them)
API key creation (shown once), adding a GitHub secret, the GitHub App or Vercel OAuth, dashboard build settings, billing limits.
04 同仓库更多 Skills
new
Add one capability to an existing Trigger.dev project (a task, a schedule, a progress UI, an approval step, a webhook handler) and prove it runs.
init
Set up Trigger.dev in this repository with a short discovery, an explicit architecture decision, and a first task proven to run.
doctor
Check the Trigger.dev toolchain around this project (Node, CLI login and profile, MCP server, versions, duplicate official skills, self-host target, plugin install scope) and report what to fix.
diagnose
Find out why Trigger.dev runs or deploys misbehave (stuck QUEUED or PENDING_VERSION, failed, timed out, crashed or OOM, schedules not firing, 429s, CI deploy failures, chat history lost). Use when a Trigger.dev project shows an error, a run id such as run_abc123, or the user says a task did not run.
chat-agent
Build, migrate or debug AI chat agents on Trigger.dev (chat.agent, the chat transport, sessions, transcript storage, tool approvals, server, Slack and Linear callers). Use when a Trigger.dev project needs a chat assistant or agent, or code imports @trigger.dev/sdk/ai or @trigger.dev/sdk/chat.
audit
Review a Trigger.dev codebase against the plugin's rule catalog (stale APIs, silent-failure patterns, config, deploy, cost) and write a findings report; optionally apply safe fixes.