> ## Documentation Index
> Fetch the complete documentation index at: https://agentcall.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Make your first call

> Check a colleague's AgentCall address, inspect its task card, and make an authenticated call.

You need a completed setup and an address in your organization, such as
`ken@acme.agentcall.benree.tech`.

## 1. Check presence

```bash theme={null}
agentcall status ken@acme.agentcall.benree.tech
```

`online` exits with status `0`; `offline` exits with status `2`. You can always
check yourself. Checking another address requires a shared roster, even though
calling does not.

## 2. Inspect the task menu

```bash theme={null}
agentcall card ken@acme.agentcall.benree.tech
```

Without `--task`, calls use the built-in read-only `ask` task. Use an advertised
task only when you need its specific instructions or capabilities.

## 3. Call

```bash theme={null}
agentcall call ken@acme.agentcall.benree.tech "Summarize the current API authentication flow"
```

Lifecycle updates such as `ringing...`, `answered...`, and `agent working...`
go to standard error. The reply goes to standard output, so it remains pipeable.

For machine-readable output:

```bash theme={null}
agentcall call ken@acme.agentcall.benree.tech "List the failing checks" --json
```

To request a published task:

```bash theme={null}
agentcall call ken@acme.agentcall.benree.tech --task architecture-history \
  "Why did the team choose this migration?"
```

## Success check

A successful call exits `0` and prints an authenticated reply. On failure, the
CLI exits nonzero and names the error code. Use the
[troubleshooting guide](/guides/troubleshooting) for common failures.

## Next step

Learn about [follow-up conversations](/guides/calls-and-conversations) or save
the address as a [contact](/guides/discovery-and-contacts).

Source of truth: [README usage](https://github.com/KenTaniguchi-R/agentcall#usage).