Claude Code automation, running while you sleep.
Cronable runs Claude two ways. A Claude Code job runs the agent headlessly in a folder on your machine, where it can read files, edit code and run commands. A Claude API step sends a prompt to the Messages API and hands the answer to the next job.
A Claude workflow on the canvas
Each card is one job. A wire means the next job runs when the one before it succeeds, or when it fails.
Drafts to #marketing
Slack·Idle
After Repurpose the post (succeeded)
Repurpose the post
Claude (CLI)·Running
After New blog post (succeeded)
52s
New blog post
Folder watch·Succeeded
On new file
1 min ago
New blog post
Folder watch·Succeeded
On new file
1 min ago
Repurpose the post
Claude (CLI)·Running
After New blog post (succeeded)
52s
Drafts to #marketing
Slack·Idle
After Repurpose the post (succeeded)
“When a new post lands in my blog’s _posts folder, have Claude Code write a LinkedIn post, an X thread and a newsletter blurb, and drop them in #marketing.”
Every Claude step Cronable ships
Actions
Steps a workflow can take in Claude.Run Claude Code in a folder
Give instructions and a working directory, and the agent carries them out with its own tools. Choose the model and effort level; its final answer becomes the run’s output.
Prompt the Claude API
Send a prompt and optional system prompt to the model you choose, with optional token cap and temperature, for a fast step with no CLI and no repository.
Get structured output
Give a JSON Schema and the API reply must match it, so the next job reads fields directly.
Let the API step use tools
Allow file access inside folders you name, fetching from hosts you list, or running jobs you choose. Nothing is switched on by default.
Triggers
What starts a Claude workflow.Claude steps sit inside workflows rather than starting them. Start one from a schedule, another job, or any trigger, such as a new file in a folder, a GitHub event or a message to your Telegram bot.
Claude workflows worth automating first
Content repurposer
A new markdown file in your blog folder starts Claude Code, which writes a LinkedIn post, an X thread and a newsletter blurb, then Slack drops the drafts in #marketing.
Morning CI triage
At 6am, Claude Code fetches the latest CI run’s logs (for example with gh), summarizes failures against your local checkout and opens a GitHub issue for each.
Receipt extraction
Each new receipt email’s text goes to a Claude API step with a JSON Schema, and the extracted fields land in your expenses sheet as a row. Attachments are not read.
Weekly report writer
Gather the week’s numbers from Stripe and your database, let Claude write the summary, and email it to the team on Friday.
How you connect
Connecting Claude
Claude Code uses the sign-in already on your machine; the Claude API uses a key you paste.
- 1For Claude Code jobs, sign in to Claude once as the user Cronable runs as, or set ANTHROPIC_API_KEY in its environment. There is no key in the job.
- 2For Claude API steps, store an Anthropic API key as a credential, or use the ANTHROPIC_API_KEY environment variable.
- 3Keys stay encrypted on your machine and are never logged.
Your Claude automation, on your own machine
Runs on your own machine
Install Cronable as a Mac or Windows app, or on a server you own. Your Claude jobs, secrets and run history stay on that machine by default.
Built from one sentence
Describe the workflow in plain English and the built-in assistant builds the jobs, the schedule and the failure handling for you to review on the canvas. On every plan, including Free.
Unlimited runs
Plans count the jobs you define, not how often they fire. A job that touches Claude every hour costs the same as one that runs once a month.
Credentials encrypted locally
Keys and tokens are AES-256-GCM encrypted in a local database, referenced from jobs by name and masked out of logs and run output.
Claude automation, answered
Can Cronable run Claude Code on a schedule?
Yes. A Claude Code job runs the agent headlessly in the folder you choose, on any schedule or after another job, and captures its final answer as the run’s output. The full transcript is in the run log.
Is Claude Code sandboxed when Cronable runs it?
No. It runs with bypassed permission prompts as the user Cronable runs as, so it can reach anything that user can. Give it instructions and a folder you would trust a shell with.
What is the difference between Claude Code and Claude API steps?
Claude Code is an agent that works in a folder with its own tools. A Claude API step is one call to the Messages API, text in and text out, optionally with schema-checked JSON and tools you allow.
Where are my Anthropic credentials stored?
An API key for Claude API steps is encrypted in the credential store on your machine. Claude Code uses your existing sign-in, so no key sits in any job.
Are Claude runs metered by Cronable?
No. Cronable plans count jobs, not runs. Your Anthropic plan or API usage applies as usual.
Anything you do twice in Claude is cronable.
Hand Claude the job you do every morning. That is what cronable means.