Cady docs

How to use the Cady CLI and the Python SDK. Each page gives the point, one example to run and what can go wrong.

The CLI is not released yet; see Install the CLI.

CLI

  1. Install the CLIOne signed script puts cady on macOS or Linux.
  2. First run: cady setupSign in, pick your company and start reporting, in one command.
  3. Check this machine: cady statusWhat this machine is attached to, and whether it will report.
  4. Choose a companySet the company per project, as a default, or per command.
  5. Call Cady from the shellRead and write company records with commands that mirror Cady’s tools.
  6. Run many calls: cady local batchRun a file of calls in order and stop at the first failure.
  7. Report usage: cady syncOne scheduled command reports this machine’s AI usage to your company.
  8. Read usage locallySee plan allowance and what ran on this machine, without sending anything.
  9. Run your own scriptsRegister a script by name, run it on a cadence and keep a journal.
  10. Upload files and export CRM dataSend a local file to Cady Files, and write CRM data to a file.

Python SDK

  1. Python SDK: your first callCall Cady from Python with CadyClient.
  2. Python SDK: errors and batchesFour errors tell you whether a call ran, and batch runs many.

Concepts

  1. Your AI app, approvals and companiesConnect Claude or ChatGPT, and how approvals apply to every call.
  2. Troubleshooting and helpExit codes, the common failures and how to reach us.

For agents: this page as Markdown.