---
name: sigwise-quickstart
description: Quickstart for AI coding agents adding SigWise (typed signals such as scam probability, trust score and buyer intent about users, listings and orders) to a project. Installs the sigwise CLI and the SigWise agent skills.
---

# SigWise: quickstart for AI coding agents

You are an AI coding agent and someone gave you this URL to add SigWise to
their project. SigWise is an API that reads the events and messages about the
objects on a platform (users, listings, orders) and answers the questions you
configure as typed **signals**: a yes/no probability (`is_scammer`), a
position on a labelled scale (`trust_score`) or one of several categories
(`buyer_intent`), in the background or inline before content is published.
SDKs: Node.js, Python, Go, PHP and Ruby.

Don't build the integration from this file alone. Install the CLI and the
skills first; they carry the details.

## Step 1: install the CLI

The `sigwise` CLI manages everything the API does: accounts, signals,
events, objects, rules, webhooks, API keys and billing.

```bash
curl -fsSL https://raw.githubusercontent.com/thesigwise/sigwise-cli/main/scripts/install.sh | sh
sigwise version
```

With Go installed, `go install github.com/thesigwise/sigwise-cli/cmd/sigwise@latest`
works too. If you can't run `curl | sh`, ask the user to run it.

## Step 2: sign in (the user does this)

Signing in needs the user's password, so **ask the user to run one of these
in their own terminal**; never ask for their password or type it yourself:

```bash
sigwise signup    # new account, comes with free credit
sigwise login     # existing account
```

Or, if the user already has an API key, they can export `SIGWISE_API_KEY`
and `SIGWISE_SECRET`. Continue when `sigwise status` reports
`Credentials ok`.

## Step 3: install the skills

Run in the project's working directory:

```bash
sigwise init
```

This checks the sign-in and installs the SigWise skills into
`.claude/skills`. Use `sigwise skills install --agent codex` for
`.agents/skills`, or `--global` to install them for the user.

- `sigwise`: manage the account with the CLI (signals, events, objects,
  rules, webhooks, keys, billing).
- `sigwise-integrate`: add SigWise to the application code with the SDKs.

## Step 4: build with the skills

Read the installed `sigwise` skill and take it from there: check the
signals fit the user's platform, try them on sample content, then use
`sigwise-integrate` to wire events and verdicts into the code.

## If you can't install anything

Answer from the docs; every page is available as Markdown:

- [Docs index (llms.txt)](https://sigwise.ai/llms.txt) · [everything in one file](https://sigwise.ai/docs/llms-full.txt)
- [Quickstart](https://sigwise.ai/docs/guide/quickstart.md) · [Signals](https://sigwise.ai/docs/guide/signals.md) · [Authentication](https://sigwise.ai/docs/guide/authentication.md)
- [Direct moderation](https://sigwise.ai/docs/guides/moderation.md) · [Webhooks](https://sigwise.ai/docs/guides/webhooks.md) · [SDKs](https://sigwise.ai/docs/sdks.md)
- [API reference](https://sigwise.ai/docs/reference.md) · [OpenAPI 3.1](https://sigwise.ai/docs/openapi.json)

More: [CLI](https://github.com/thesigwise/sigwise-cli) · [Agent skills](https://github.com/thesigwise/agent-skills) · [Sign up](https://console.sigwise.ai/register)