← Back to Skills Marketplace
50
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install okx-cex-auth
Description
Use this skill when the user wants to 'login/log in/sign in', 'authenticate', 'authorize', 'connect OKX account', 'set up credentials', 'first time setup', '...
Usage Guidance
This skill appears to implement an OKX CLI login flow and mostly behaves as expected, but take these precautions before installing or using it:
- Confirm the package identity before running npm install: check the npm publisher for @okx_ai/okx-trade-cli, verify the package homepage/maintainer, and prefer installing only from trusted sources.
- Resolve the metadata mismatch: the skill's SKILL.md requires the `okx` binary and offers an npm install, but the registry metadata did not list those requirements — ask the publisher or registry maintainer to fix the manifest.
- Avoid pasting API keys/secrets into the chat. If the skill asks you to provide AK/SK/PP, prefer to run the CLI locally yourself (npm install locally) and perform `okx config init` rather than sending credentials to the agent.
- If you must use the agent to configure auth, limit the scope: create API keys with minimal permissions and rotate them after use.
- Inspect `~/.okx/config.toml` and backup any existing config before the skill modifies or deletes profiles.
If you want higher assurance, request the skill author/publisher to provide signed release artifacts or a clear, matching registry manifest showing the required binaries and install steps.
Capability Analysis
Type: OpenClaw Skill
Name: okx-cex-auth
Version: 1.3.2
The skill bundle provides structured instructions for an AI agent to handle OKX exchange authentication via the `@okx_ai/okx-trade-cli` tool. It implements a standard OAuth 2.0 device flow and includes rigorous 'Pre-flight checks' (Step 0) to ensure correct regional site selection (Global, EEA, or US) and to prevent conflicts with existing API keys. The instructions are designed to be non-blocking and transparent, requiring the agent to surface verification URLs directly to the user and wait for manual confirmation rather than auto-polling. No indicators of data exfiltration, malicious command execution, or unauthorized persistence were found.
Capability Tags
Capability Assessment
Purpose & Capability
The SKILL.md behavior (running `okx` commands, guiding OAuth/device flow, handling API keys) aligns with an OKX-auth helper. However the top-level registry metadata provided to the scanner lists no required binaries/install, while the SKILL.md declares a required `okx` binary and an npm install of @okx_ai/okx-trade-cli — this metadata mismatch is an inconsistency that should be resolved.
Instruction Scope
Runtime instructions are narrowly focused on auth: running `okx config show --json`, `okx auth status --json`, parsing `~/.okx/config.toml` semantics, performing `okx auth login`, and guiding the user through API-key vs OAuth choices. The skill does instruct the agent to ask users for API credentials if needed, which is within the scope of an auth helper but is a sensitive operation.
Install Mechanism
No install spec in the registry metadata, but SKILL.md recommends installing `@okx_ai/okx-trade-cli` via npm (global install). Installing a package from the public npm registry is expected for a CLI, but it carries moderate supply-chain risk — verify the package name, publisher, and consider installing locally or inspecting the package before running globally.
Credentials
The skill does not declare any environment-variable or config path requirements beyond interacting with the OKX CLI and `~/.okx/config.toml`. However it explicitly instructs agents to obtain API Key/Secret/Passphrase from the user (or guide them to run `okx config init`), which is proportionate to the purpose but sensitive: the agent/skill (and platform) may receive secrets if users paste them into chat. The skill should clearly instruct users not to share secrets in public contexts and prefer local CLI usage.
Persistence & Privilege
The skill is not always-enabled, does not request system-wide privileges, and does not modify other skills. It does instruct modifications to the user's OKX config (adding/removing profiles) which is appropriate for an auth tool and within its scope.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install okx-cex-auth - After installation, invoke the skill by name or use
/okx-cex-auth - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.3.2
okx-cex-auth v1.3.2
- Updated documentation with stricter pre-login checks, including a detailed Step 0 pre-flight check to avoid defaulting user region.
- Clarified handling of site selection: the site must be explicitly chosen by the user before login, never defaulted.
- Documented requirements for handling pre-existing API key profiles, including how to address invalid API keys.
- Expanded and enforced rules for OAuth flow initiation, especially around user interaction for regional site choice.
- Provided clearer guidelines on when and how the site field should be obtained and used for login.
- Added comprehensive explanation of agent responsibilities before starting any authentication flow.
Metadata
Frequently Asked Questions
What is okx-cex-auth?
Use this skill when the user wants to 'login/log in/sign in', 'authenticate', 'authorize', 'connect OKX account', 'set up credentials', 'first time setup', '... It is an AI Agent Skill for Claude Code / OpenClaw, with 50 downloads so far.
How do I install okx-cex-auth?
Run "/install okx-cex-auth" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is okx-cex-auth free?
Yes, okx-cex-auth is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does okx-cex-auth support?
okx-cex-auth is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created okx-cex-auth?
It is built and maintained by numpy0001 (@numpy0001); the current version is v1.3.2.
More Skills