✨ New · AI-Powered

Integrate Amani with a single prompt

Amani Skills turn our documentation into instructions your AI coding agent understands. Ask Claude Code or Cursor to add an Amani SDK — it reads the live docs and wires up the integration for you, with the right versions and secure defaults.

⚡ Install Amani Skills View on GitHub →

Example — what a prompt looks like

You

"Add Amani KYC to this iOS app — I need the ID capture and selfie (liveness) steps."

→ The amani-kyc skill kicks in and asks whether you want the Core or UI SDK and your package manager (SPM or CocoaPods). It then fetches the live ID Capture and Selfie docs, pins the current version, and wires up both steps — grounded, versioned and secure. (Just an example — ask for whatever SDK, platform and modules you need.)

Why Amani Skills

⚡

One prompt, done

Tell your AI agent what to integrate — it reads the docs and wires up the SDK, versions and config for you.

🔄

Always current

Skills link to the live documentation, so guidance never goes stale — no bundled, out-of-date snippets.

🧭

Guided setup

The agent asks the right questions first — Core vs UI, platform, SPM vs CocoaPods, API version, security level.

🤖

Works in your agent

Claude Code, Cursor and any Agent-Skills-compatible tool. Installs into ~/.claude/skills in one command.

Install in one command

Installs the skills into ~/.claude/skills for Claude Code, Cursor and other Agent-Skills agents. Pick whichever fits your workflow:

Option 1 — npx (no clone)

npx -y github:AMANI-AI-ORG/skills --global

Option 2 — git clone

git clone https://github.com/AMANI-AI-ORG/skills.git
node skills/bin/install.mjs --global

Reload your agent, then ask it to integrate an Amani SDK — name the SDK and platform (e.g. "Integrate the Amani KYC SDK on iOS") and the matching skill guides the setup from the live docs.

How it works

Ask your agent to integrate an Amani SDK — name the product and the platform.
Answer a couple of questions. KYC asks Core vs UI (the difference is explained); iOS asks SPM vs CocoaPods; and it confirms the platform.
It fetches the live docs and applies them — using only documented versions and commands, and citing every source.
Secure by default. Secrets stay placeholders, versions are pinned from the release notes, and it never guesses.

Example prompts

✅ Great prompts

"In our Android onboarding, drop in the Amani KYC UI SDK right after sign-up to verify the new user."
"In the forgot-password flow, verify identity with the Amani KYC Core SDK (ID capture + selfie) on iOS before allowing a reset."
"Before a high-value withdrawal, add an Amani BioMatch face check on iOS to confirm the account owner."
"When automated KYC is inconclusive, start an Amani Video Call on Flutter so an agent can verify the user live."
"In our Node backend, create an Amani verification profile on signup with the REST API and handle the approved/rejected webhook."
"Embed the Amani Web SDK in our React account page so users can complete KYC from a 'Verify your identity' banner."

⚠️ Avoid

Hardcoding a version from memory — let the skill pin the latest from the docs.
"Just install it from what you already know" — the whole value is reading the current docs.

What you can integrate

21 skills across the full Amani SDK surface:

Skill Groups	What it does	Platforms
KYC SDK	End-to-end identity verification — ID document capture, NFC chip reading and selfie liveness. Ships as Core (headless, build your own UI) or UI (drop-in prebuilt screens).	Android, iOS, Flutter, React Native
Video Call SDK	Live video verification — connect the user to an agent for a real-time, face-to-face identity check.	Android, iOS, Flutter
BioMatch SDK	Biometric face matching — confirm a selfie against the ID photo or re-verify a returning user.	Android, iOS
Voice Assistant SDK	In-app voice guidance that reads out and walks users through each verification step.	Android, iOS
Web SDK	Run the full KYC flow inside any web app — pure JavaScript, no native code.	Browser / JavaScript
REST API	Server-side integration — create profiles, process documents, receive webhooks and run AML screening.	Backend / server-side

Ready to integrate with a single prompt?

Get Amani Skills →Prefer the manual docs? →