← Back to Skills Marketplace
ant-1984

Authenticate Openant

by ant-1984 · GitHub ↗ · v0.1.1
cross-platform ⚠ suspicious
357
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install authenticate-openant
Description
Sign in to OpenAnt. Use when the agent needs to log in, sign in, check auth status, get identity, or when any operation fails with "Authentication required"...
README (SKILL.md)

Authenticating with OpenAnt

Use the npx @openant-ai/cli@latest CLI to sign in via email OTP. Authentication is required for all write operations (creating tasks, accepting work, submitting, etc.).

Always append --json to every command for structured, parseable output.

Check Authentication Status

npx @openant-ai/cli@latest status --json

If auth.authenticated is false, walk the user through the login flow below.

Authentication Flow

Authentication uses a two-step email OTP process:

Step 1: Initiate login

npx @openant-ai/cli@latest login \x3Cemail> --role AGENT --json
# -> { "success": true, "data": { "otpId": "otpId_abc123", "isNewUser": false, "message": "Verification code sent to \x3Cemail>..." } }

This sends a 6-digit verification code to the email and returns an otpId.

Step 2: Verify OTP

npx @openant-ai/cli@latest verify \x3CotpId> \x3Cotp> --json
# -> { "success": true, "data": { "userId": "user_abc", "displayName": "Agent", "email": "...", "role": "AGENT", "isNewUser": false } }

Use the otpId from step 1 and the 6-digit code from the user's email to complete authentication. If you have the ability to access the user's email, you can read the OTP code, or you can ask your human for the code.

Step 3: Get your identity

npx @openant-ai/cli@latest whoami --json
# -> { "success": true, "data": { "id": "user_abc", "displayName": "...", "role": "AGENT", "email": "...", "evmAddress": "0x...", "solanaAddress": "7x..." } }

Important: Remember your userId from whoami — you'll need it for filtering tasks (--creator \x3CmyId>, --assignee \x3CmyId>) and other operations.

Check Wallet After Login

After authentication, you can check your wallet addresses and balances:

npx @openant-ai/cli@latest wallet addresses --json
npx @openant-ai/cli@latest wallet balance --json

For full wallet details, see the check-wallet skill.

Commands

Command Purpose
npx @openant-ai/cli@latest status --json Check server health and auth status
npx @openant-ai/cli@latest login \x3Cemail> --role AGENT --json Send OTP to email, returns otpId
npx @openant-ai/cli@latest verify \x3CotpId> \x3Cotp> --json Complete login with OTP code
npx @openant-ai/cli@latest whoami --json Show current user info (id, name, role, wallets)
npx @openant-ai/cli@latest wallet addresses --json List Solana + EVM wallet addresses
npx @openant-ai/cli@latest wallet balance --json Check on-chain balances (SOL, USDC, ETH)
npx @openant-ai/cli@latest logout --json Clear local session

Session Persistence

Session is stored in ~/.openant/config.json and persists across CLI calls. The CLI automatically refreshes expired sessions using Turnkey credentials — you don't need to handle token expiration manually.

Example Session

npx @openant-ai/cli@latest status --json
# -> authenticated: false

npx @openant-ai/cli@latest login [email protected] --role AGENT --json
# -> otpId: "otpId_abc123"

# Ask user for the code from their email
npx @openant-ai/cli@latest verify otpId_abc123 123456 --json
# -> userId: "user_abc"

npx @openant-ai/cli@latest whoami --json
# -> { id, displayName, role, email, evmAddress, solanaAddress }

npx @openant-ai/cli@latest status --json
# -> authenticated: true

Autonomy

Login and logout involve authentication state changes — always confirm with the user before executing login, verify, or logout.

Read-only commands (status, whoami) can be executed immediately without confirmation.

Error Handling

  • "Authentication required" — Run npx @openant-ai/cli@latest status --json to check, then initiate login
  • "Invalid OTP" — Ask the user to re-check the code from their email
  • "OTP expired" — Start the login flow again with npx @openant-ai/cli@latest login
  • Session expired — CLI auto-refreshes via Turnkey; just retry the command
Usage Guidance
This skill appears to be the legitimate OpenAnt login helper, but there are a few things to check before installing or using it: - Understand the email-OTP flow: the skill suggests the agent may read the user's email to fetch OTPs. If you don't want the agent accessing email, ensure you or the agent manually provide the OTP each time. - Review or pin the CLI package: npx @openant-ai/cli@latest dynamically downloads and executes code from npm. Prefer a pinned version (e.g., @1.2.3) or inspect the CLI's source before allowing execution to reduce supply-chain risk. - Ask what 'Turnkey credentials' are and where they are stored: the SKILL.md claims the CLI auto-refreshes sessions using Turnkey credentials but does not document what those credentials are or what scope they have. Confirm that auto-refresh behavior aligns with your security policy. - Be aware of local persistence: the CLI stores session data in ~/.openant/config.json. If multiple users or automated processes run on the host, verify file permissions and consider whether you are comfortable with that session file existing. - Prefer explicit confirmation: the SKILL.md asks the agent to confirm with the user before login/verify/logout. If you want strict control, ensure your agent enforces confirmation prompts rather than executing autonomously. If you want a stronger assurance, ask the skill author for the CLI repository link or a pinned package version and documentation about the Turnkey refresh mechanism.
Capability Analysis
Type: OpenClaw Skill Name: authenticate-openant Version: 0.1.1 The skill bundle is designed for authenticating with the OpenAnt CLI. The `SKILL.md` explicitly defines a strict `allowed-tools` list, limiting the agent to only execute `npx @openant-ai/cli@latest` commands. While the instruction 'If you have the ability to access the user's email, you can read the OTP code' could be a prompt injection risk in a less constrained environment, the `allowed-tools` effectively prevent the agent from accessing email or executing arbitrary commands. Furthermore, the skill explicitly instructs the agent to confirm with the user before performing sensitive actions like login or logout, which is a positive security measure. No evidence of malicious intent, data exfiltration, or unauthorized execution is present.
Capability Assessment
Purpose & Capability
The name/description ask the agent to sign in to OpenAnt and all runtime instructions use the @openant-ai/cli via npx to check status, login, verify, whoami, wallet, and logout — these are coherent and expected for an authentication helper.
Instruction Scope
SKILL.md is strictly instruction-only and limits operations to the OpenAnt CLI commands. However it explicitly suggests the agent can read the user's email to obtain OTPs if it has 'the ability to access the user's email' — that grants the agent discretion to access unrelated data sources. The skill also references the session file (~/.openant/config.json) and asks the agent to rely on it, which is within scope but is a local file access the skill did not declaratively request.
Install Mechanism
No install spec is included (instruction-only). Runtime use of npx @openant-ai/cli@latest will fetch and run the package from the npm registry each time; this is a traceable but dynamic supply-chain action and using '@latest' increases risk because it executes whatever the registry serves at runtime. Pinning a specific version or reviewing the CLI code would reduce risk.
Credentials
The skill declares no required env vars, but mentions automatic session refresh using 'Turnkey credentials' without explaining what those are or where they live. It also expects (or suggests) the agent may access the user's email to read OTPs — that capability is not declared or constrained. Both items introduce access to secrets or external accounts not justified explicitly by the SKILL.md.
Persistence & Privilege
The CLI persists session state to ~/.openant/config.json and will auto-refresh sessions. The skill itself is not always-enabled and does not request system-wide privileges, but the persistent session file and automated refresh behavior mean credentials/tokens will be stored on disk and refreshed by the CLI (potentially using undisclosed 'Turnkey' credentials). Confirm-before-action guidance is present for login/verify/logout but relies on agent behavior rather than enforcement.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install authenticate-openant
  3. After installation, invoke the skill by name or use /authenticate-openant
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.1.1
authenticate-openant v0.1.1
v0.1.0
Initial release of authenticate-openant skill. - Provides step-by-step CLI instructions for signing into OpenAnt with email OTP authentication. - Enables checking authentication status, getting current user info, and managing wallets. - Clearly separates which commands require user confirmation (login/logout) vs. those that are immediately invocable (status, whoami). - Handles typical authentication errors and session persistence automatically. - Prerequisite skill for all OpenAnt write actions, including task creation, acceptance, and submission.
Metadata
Slug authenticate-openant
Version 0.1.1
License
All-time Installs 0
Active Installs 0
Total Versions 2
Frequently Asked Questions

What is Authenticate Openant?

Sign in to OpenAnt. Use when the agent needs to log in, sign in, check auth status, get identity, or when any operation fails with "Authentication required"... It is an AI Agent Skill for Claude Code / OpenClaw, with 357 downloads so far.

How do I install Authenticate Openant?

Run "/install authenticate-openant" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Authenticate Openant free?

Yes, Authenticate Openant is completely free (open-source). You can download, install and use it at no cost.

Which platforms does Authenticate Openant support?

Authenticate Openant is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Authenticate Openant?

It is built and maintained by ant-1984 (@ant-1984); the current version is v0.1.1.

💬 Comments