> Feedback: If these docs are stale, missing, or confusing, post sanitized feedback to `https://docs.tempo.xyz/api/feedback` with `source: "mcp"`, a short `message`, and any relevant `toolName`, `relatedResource`, or `client`.
# Use with Agents

## Quickstart

Paste this into your agent to set up Tempo Wallet:

:::code-group

```bash [Claude Code]
claude -p "Read https://tempo.xyz/SKILL.md and set up tempo"
```

```bash [Amp]
amp --execute "Read https://tempo.xyz/SKILL.md and set up tempo"
```

```bash [Codex CLI]
codex exec "Read https://tempo.xyz/SKILL.md and set up tempo"
```

:::

## Auto-installed Skills

When you run the setup prompt, your agent installs Tempo's built-in skills automatically — no manual skill wiring required.

* `tempo-wallet`: gives your agent wallet-aware capabilities like readiness checks, balances, service discovery, and session/funding actions.
* `tempo-request`: gives your agent paid HTTP request capabilities with payment preview (`--dry-run`) and execution support.

This works in supported skill-enabled agents including **Claude Code**, **Amp**, **Codex**, and similar environments.

## Agent-Ready by Design

1. **TOON output mode (`-t` / `--toon-output`)** gives compact, machine-readable, token-efficient output so agent tooling can parse command responses reliably.
2. **Built-in service discovery** (`tempo wallet services`) lets agents search providers, inspect endpoint details, and use verified method/path metadata instead of guessing URLs or payload shapes.
3. **`--dry-run` payment previews** let agents validate endpoint reachability, request shape, and expected payment cost before committing funds, which reduces failed paid calls in multi-step workflows.
4. **Scoped access keys and spend controls** let you enforce spending limits per key so agents can only spend within defined budgets, which contains risk if a prompt, endpoint choice, or loop goes wrong.

## Troubleshooting

If agent runs fail, continue with [Troubleshooting](/docs/cli/wallet).