What does the developer actually see?

A short question in the terminal or the SDK error UI. Not a chatbot landing page. They can answer in a sentence. If they walk away, you still have the stack, the OS, and the command they ran.

Do tickets go out on their own?

No. AethelLayer drafts GitHub and Linear. You approve. Same rule as the rest of how we work: agents propose, humans ship.

Hashed identifiers, runtime, OS, the error, the thread, and optional session links. Not source code dumps. Not secrets you did not already print to the terminal.

How fast is a real install?

Most teams are live before lunch. The 60-day pilot is for teams who want us in the repo, writing the taxonomy with you, not a slide about it.

AeAethelLayer

AethelLayer

Catch the failure before they leave.

A probe in your CLI, SDK, and API that keeps the stack and opens the thread.

Paste it in a terminal. It wraps a command that exits 1 and prints a drafted GitHub issue. No account.

Early access Talk to sales How a probe becomes a ticket

payments-api · zshNode v20.10 · Darwin 24.3

developer@acme payments-api %

$ npx @acme/cli auth login --env staging

→ pulling compose overlay · auth-worker

→ exchanging OAuth code at /oauth/callback

ERR_DOCKER_OOM container auth-worker killed (exit 137)

AETHELLAYER MICRO-PROBE212ms

Caught ERR_DOCKER_OOM on POST /oauth/callback. auth-worker is 512Mi. Node asked for a 4GB heap.

What were you trying to do? I will attach it to the GitHub issue.

CLI login inside Compose for a staging cut

stack_traceexit 137

Fatal process out of memory: Heap
  at JWT.exchange (oauth.js:88)
  at AuthorizeHandler (payments.go:64)
  container: auth-worker
  limit: 512Mi  rss: 498Mi

GHacme/payments-apidraft

#482 ERR_DOCKER_OOM during OAuth callback

Raise auth-worker to 1Gi. You still hit send.

Docker OOMOAuth callbacksgRPC deadlinesK8s CrashLoopJWT expiryRate limitsCLI onboardingSDK install frictionAPI middlewareLinear auto-draftGitHub issuesPostHog drop-offDocker OOMOAuth callbacksgRPC deadlinesK8s CrashLoopJWT expiryRate limitsCLI onboardingSDK install frictionAPI middlewareLinear auto-draftGitHub issuesPostHog drop-off

of developers never file a ticket. They just leave.

minutes is usually enough to put the probe in a real CLI

places it lives: CLI, SDK, and API middleware

Why this keeps happening

They were not angry. They just did not want to write a ticket.

Nine out of ten developers never file anything. They type the command, it dies, and they move on. Your charts still look fine.

quiet churn · 10 evaluators

filed a ticket

left in silence

toolgives youstill missing

Datadogexit 137why they left

NPS surveya score on Fridaythe error itself

AethelLayerstack + intent + draftnothing you care about

How it works

Add it once. Catch the moment. Ship the ticket.

The work shows up while the error is still on their screen.

SDK dies

OOM, JWT, gRPC

→

Probe keeps it

stack + OS + cmd

→

One question

in the terminal

→

Draft ticket

you hit send

Drop in the probe

One line in the CLI, the Node SDK, or API middleware. About ten minutes. Read-only to your product.

$ npm i @aethellayer/sdk
probe.wrap(cli)

They hit a real error

Docker OOM. A JWT expires. gRPC dies. The probe already has the stack. Then it asks one short question.

ERR_DOCKER_OOM auth-worker killed (exit 137)

You get a thread, not a guess

An inbox that feels like Intercom, for developers. A GitHub draft sits next to it. Nothing publishes until you say yes.

Draft #482 waiting · you still hit send

See the full walkthrough →

Product

The layer that sits in the failure, not in a form you send later.

CLI, SDK, and API. One probe. One inbox. A ticket you still approve.

CLI

$ npx @acme/cli login
ERR_DOCKER_OOM
probe: what were you doing?

SDK

probe.wrap(client)
JWT expired at +46s
callback never returned

API

POST /oauth/callback
gRPC deadline exceeded
draft #482 waiting

A developer inbox

Threads look like Intercom. The people in them are developers who just hit your error, not a chatbot landing page.

ERR_DOCKER_OOM

auth-worker

JWT_EXPIRED

auth-worker

gRPC deadline

auth-worker

live · 212ms

Caught OOM on /oauth/callback

CLI login inside Compose for a staging cut

Draft GitHub #482 · waiting on you

A ticket that waits for you

GitHub and Linear drafts in the language your eng team already uses. PostHog on the side if you wired it. Nothing ships until you say yes.

GitHub issue #482 drafted

Linear ticket ready

PostHog session attached

Explore the product →

Live demo

Watch it happen in the terminal, not in a slide.

Run the failing login. See the probe catch ERR_DOCKER_OOM. Type what you were actually doing. A GitHub issue draft appears. That is the whole product, in one go.

aethel · probe · staging

developer@acme payments-api %

Replay the moment a developer hits Docker or OAuth friction. The probe fires inside the terminal, keeps the root cause, and drafts the ticket for someone on your team to approve.

After the first week

What platform teams say when the quiet ones finally show up

“We thought onboarding was fine because nobody complained. The first week of probes showed three Docker OOMs on the exact page we call the happy path. That is not feedback. That is the product talking.”
Head of DX
Payments SDK, 40-person platform team
First week covered the quarter

“I asked a developer what they were trying to do. They typed one sentence. We had a Linear ticket with the stack, the Node version, and the compose file before they closed the tab.”
Founder
Devtools, Series A
Ticket in the same minute

“PostHog said they dropped at /oauth/callback. AethelLayer told us why: the JWT expired 46 seconds before the popup returned. We stopped guessing.”
VP Engineering
Auth platform, 120 employees
Root cause, not a chart

Straight answers

The questions we get on the first call

No. You wrap the CLI, hang middleware on the API, or register one file in Next.js. The probe watches failures. It does not sit in your hot path as a new product surface, and it does not ask you to change how developers already use your tool.

Early access

Stop finding out about churn in a dashboard three days later.

Paste npx --yes @aethellayer/sdk run -- false and read the drafted issue in your terminal. Book a call when you want the probe in your CLI.

Early access Talk to sales