/Docs

CLI reference

.md

slasha deploy & logs

slasha deploy starts a deployment. slasha logs shows a deployment’s build and runtime logs, and can follow them live.

slasha deploy

Builds and deploys a commit that’s already in the app’s repository on the server.

slasha deploy [--commit <SHA>] [-f] [--app <slug>]
Flag Description Default
--commit <SHA> Full 40-character commit SHA to deploy. It must already be in the app’s repository on the server. Latest commit on the app’s default branch
-f, --follow After triggering, stream the deployment’s logs Off
# Deploy the latest commit on main
slasha deploy

# Deploy and watch the build; exits 1 if the deployment fails
slasha deploy -f

# Deploy an exact commit
slasha deploy --commit "$(git rev-parse HEAD)"

Without -f, it returns once the deployment is queued:

SUCCESS: Deployment triggered.
  Commit: 9f1c2b7e4a0d5c8b3e6f1a2d4c7b9e0f1a3c5d7e

Follow logs: slasha logs --app storefront <deployment-id> --follow

For what happens after that, see How it works and Health checks.

Notes:

  • It deploys what the server has, not your working tree. For a Slasha Git app, git push first. An HTTPS push with Auto-Deploy on already deploys, so you need slasha deploy only to redeploy, deploy after an SSH push, or pick a commit.
  • GitHub and Git URL apps fetch the latest code before deploying.
  • Short SHAs aren’t resolved. Use git rev-parse HEAD.
  • An app runs one operation at a time. If a deploy, scale or restart is in progress, the command fails and says what the app is doing.
  • With -f, a failed deployment exits 1 with Deployment <id> failed. A successful one keeps following the runtime logs until Ctrl-C. In CI, leave out -f or add a timeout.
  • To cancel a build, use the dashboard.

slasha logs

Shows one deployment’s logs: build, release phase, and output from its web and worker containers.

slasha logs [ID] [--deployment-id <ID>] [-f] [-n <N>] [--search <TEXT>] [-p <PREFIX>] [-s <STREAM>] [--app <slug>]
Argument / flag Description Default
ID Deployment ID (from slasha deployments list) Newest deployment, whatever its status
--deployment-id <ID> Same as ID, as a flag. Can’t combine with ID.
-f, --follow Print recent lines, then keep streaming new ones Off
-n, --limit <N> How many of the most recent lines to fetch 2000
--search <TEXT> Only lines whose message contains TEXT (case-insensitive)
-p, --prefix <PREFIX> Only lines from one source, e.g. system, web, web.0, worker.1
-s, --stream <STREAM> Only stdout or stderr Both

slasha deployments logs is the same command. Each line looks like:

2026-11-01 14:03:12 [web.0] [stdout] Listening on port 8080

Local time, prefix (system for Slasha’s own messages, web.0, worker.1… for containers), stream, message.

Examples

# Latest deployment, opened in a pager (q to quit, / to search)
slasha logs

# Follow the latest deployment live
slasha logs -f

# Only errors from the web containers
slasha logs -p web -s stderr

# Search the last 500 lines for "timeout"
slasha logs -n 500 --search timeout

# A specific deployment, piped to grep (no pager when piped)
slasha logs 3b9e6f0c-2a41-4d7e-9c55-8f1a2b3c4d5e | grep -i migration

How following works

  • -f prints the last --limit matching lines, then streams new ones.
  • It keeps going while the deployment builds or runs, and stops on its own when it ends up failed or stopped (e.g. replaced by a newer one), printing Deployment is failed. or Deployment is stopped.
  • If it had already failed or stopped, you get its history and nothing more to follow.
  • Ctrl-C stops it any time.

Notes:

  • Without ID you get the newest deployment, even a failed one. After a failed deploy, pass the running deployment’s ID to see its logs.
  • -p web matches every web container. -p web.1 matches only that one.
  • Logs are kept until the deployment or app is deleted.
  • For database logs, use slasha services logs.