Skip to content
Scalpels home

Scalpels is officially launching on October 31, 2026. Special early adopter pricing is available now.

Wake

Laravel MIT licensed Experimental

Point every Laravel application you run at one Wake installation and the log lines land as Parquet on a bucket you already pay for, with the catalog in your own Postgres. There is no log viewer to log into. Your agent reads across all of your apps through a single MCP server, which is the only way anybody looks at logs here.

Deploy Wake as a supporter Or run it yourself, free Read the Docs

Work in progress. Available to supporters.

Replaces
Hosted log-management subscriptions
Runs on
Your Laravel Cloud account
Needs
Compute · Database · Queue · Object storage · Scheduler

What it keeps

Parquet on your bucket, catalog in your Postgres, read through one MCP server

  • One channel per app

    Add `wake` to a Laravel app's log stack and every line it writes ships to your installation.

  • The whole log line

    Message, level, channel, context, extra, exception with a capped trace, environment, hostname and a correlation id.

  • Search and tail

    Text or regex over a window, newest first, or the most recent lines since a cursor.

  • Errors across the fleet

    The last errors for one app or all of them, answered from a table built for the long range.

  • The lines around one event

    Hand an agent an event id and it reads what happened either side of it in that app.

  • Ingest health

    Drops, rejections, rate-limit hits, flush lag and which client version each app is on.

  • No viewer, on purpose

    MCP

    Wake ships ten MCP tools and no log search screen. People use the admin pages to add an app and hold its credential; the logs themselves are only ever read by an agent, which is what a fleet of forty applications actually needs — nobody reads forty log streams, and an agent can be asked "which apps are throwing this exception?" in one question. Every response carrying log content is marked untrusted, because a log line is data somebody else wrote.

  • Raw by default, and it says so

    The client sends your log payloads unchanged. That is the right default for debugging and the wrong one for a context array holding a token, so the client calls a payload filter on every event: rebind it in your own service provider to redact a value or return null to drop the event entirely. Wake will not quietly scrub anything on your behalf, and the docs say which risk you are taking.

Wake vs hosted log management

One slice of their product, on your own bucket

Papertrail, Better Stack and Datadog run large, polished log platforms with alerting, dashboards and collectors for everything. Wake does a narrow thing: take Laravel log lines from many apps, keep them cheaply, and let an agent read across them. The question is not which is better but whether the parts you use are the parts you are paying for.

Wake feature comparison
Feature Wake Self-hosted, MIT, Experimental Papertrail Hosted, per GB ingested Better Stack Hosted, SQL and an MCP server Datadog Logs Hosted, per-index retention
Many apps, one searchable sink the row we rebuilt Does this well:Your bucket Does this well:Included Does this well:Included Does this well:Included
An agent reads your logs the row we rebuilt Does this well:Ten MCP tools Not built:REST API only Does this well:MCP server Partly, or with a caveat:API and add-ons
Retention you simply choose the row we rebuilt Does this well:A number you set Partly, or with a caveat:Archive to S3 Does this well:Per source Works, but on their terms:Per index, per contract
A log viewer people sit in we did not build this Not built:No viewer at all Does this well:Live tail UI Does this well:Polished Does this well:Industry standard
Alerting, anomalies, on-call we did not build this Not built:Not built Does this well:Included Does this well:Anomaly detection Does this well:Extensive
Metrics, traces, RUM alongside we did not build this Not built:Logs only Not built:Logs only Does this well:Full suite Does this well:All of it
Collectors for anything Not built:Laravel client only Does this well:Syslog and HTTP Does this well:eBPF, Vector, OTel Does this well:Everything
Redaction before storage Partly, or with a caveat:You write the filter Partly, or with a caveat:Filter on ingest Does this well:Built in Does this well:Sensitive data scanner
Where your log lines live Does this well:Your bucket Works, but on their terms:Their cluster Works, but on their terms:Their cluster Works, but on their terms:Their cluster
Cost of the next million lines Does this well:About 56 MB Works, but on their terms:Per GB ingested Works, but on their terms:Per plan and usage Works, but on their terms:Per index, per GB
Who is on call for the sink Partly, or with a caveat:You are Does this well:They are Does this well:They are Does this well:They are
  • Does this well
  • Partly, or with a caveat
  • Not built
  • Works, but on their terms

The honest test

Open last month's log bill and ask who actually used it. Moving loses little when these are true.

  • Everything that sends logs is a Laravel application.
  • When something breaks, an agent reads the logs before a person does.
  • You never set up a log-based alert, because the alerts you have come from somewhere else.

Then you are paying per gigabyte for storage you own, and the same slice runs on a bucket and a Postgres you are already billed for. Thirty days of retention is a number you set rather than a plan you upgrade.

Keep paying them if

These are the parts we chose not to build, and we are not going to pretend otherwise.

  • Somebody needs a log screen to sit in and watch.
  • Log patterns page an on-call engineer, and that matters.
  • Your estate is polyglot and the logs have to arrive from everything.
  • You want logs, metrics and traces correlated in one product.

Small, sharp tools are a deliberate scope, not an unfinished roadmap. Wake will not grow into an observability platform.

Be honest about it

You don't need us to run Wake

The repository is public and MIT licensed. These are the steps, and you are welcome to spend the afternoon doing them by hand.

  1. 1

    Fork the Wake repository into your own GitHub organisation.

  2. 2

    Deploy it through Scalpels with Postgres, a 2 GiB managed queue, a bucket and the scheduler.

  3. 3

    Connect each application from Scalpels, which mints its credential, writes WAKE_URL and WAKE_TOKEN, and opens the client-install pull request. Add `wake` to that application's log stack and it reports.

Or become a supporter, and Scalpels deploys it for you. It forks it, provisions exactly what it needs on your Laravel Cloud account, and deploys it. The app and the infrastructure are yours either way — the only question is who spends the afternoon.

Deploy Wake as a supporter

Guides for Wake

Want some help?

After you buy, we will pay for an hour with one of these experts to help you get everything up and running.

  • Artisan Build

    USA

    We run Scalpels, as well as a suite of other tools for developers and development teams.

    Agency Ask for help
    Ask Artisan Build for help

Wake pairs with the rest of the suite

Hone watches how fast those same applications are and Telltale watches what people do in them, so between the three the question "what happened?" has somewhere to go whatever shape it arrives in. Capstan gives the agent reading all of it a workspace of its own.

Hone Telltale Capstan