Claude Code CLI for OpenClaw
v1.0.2Install, authenticate, and use Claude Code CLI as a native coding tool for any OpenClaw agent system.
⭐ 0· 400·1 current·1 all-time
byAsif@asif2bd
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name/description claim to install and configure the Claude Code CLI; the package only requires node/npm and the script installs @anthropic-ai/claude-code via npm and guides OAuth setup. Required binaries and content align with the stated purpose.
Instruction Scope
SKILL.md and install.sh instruct the user to run npm install -g @anthropic-ai/claude-code, run interactive 'claude setup-token', and persist the resulting OAuth token in shell RC or OpenClaw config. This is within the skill's scope, but it includes a recommendation to add the token to OpenClaw config which will make it available to agents — intended for use but something users should consciously accept. The README references MissionDeck.ai session tracking; that appears informational (no code in the repo performs network calls to MissionDeck).
Install Mechanism
There is no network-download of arbitrary archives; the installer uses npm to install the official @anthropic-ai/claude-code package. That is a standard public-registry install and proportionate to the task.
Credentials
The only real secret the skill explains is the Claude OAuth token (CLAUDE_CODE_OAUTH_TOKEN), which is necessary for the described OAuth-based workflow. The skill does not request unrelated credentials. Users should be aware that persisting the token into OpenClaw config makes it accessible to agent processes that use that backend.
Persistence & Privilege
The skill does not demand always-enabled privilege, does not modify other skills, and only writes to the user's shell RC files and suggests a user-modified OpenClaw config.patch. Those are normal installer behaviors and are scoped to the user's environment.
Assessment
This package is largely documentation plus an installer that runs 'npm install -g @anthropic-ai/claude-code' and guides you through the official Claude OAuth flow. Before installing: 1) verify the npm package @anthropic-ai/claude-code is the official Anthropic package (check its npm and GitHub pages), 2) be aware that adding the OAuth token to ~/.openclaw/config.patch or an env file will make that token usable by agent backends — only do this if you trust the agents and host, 3) prefer secrets managers or OpenClaw's secret storage for production rather than putting long-lived tokens in plaintext files like /etc/environment or ~/.bashrc, 4) the OAuth setup requires an interactive TTY (claude setup-token), and 5) review the installed 'claude' binary location and run 'claude --version' and check the package source if you need extra assurance. The skill's references to MissionDeck.ai are informational; if you enable any cloud/dashboard integration, review what data is sent and where.Like a lobster shell, security has layers — review code before you run it.
agent-toolsvk973qbr2m4b7gxn4ey3q4eyg3n82tckcanthropicvk973qbr2m4b7gxn4ey3q4eyg3n82tckcclaude-codevk973qbr2m4b7gxn4ey3q4eyg3n82tckcclivk973qbr2m4b7gxn4ey3q4eyg3n82tckccodingvk973qbr2m4b7gxn4ey3q4eyg3n82tckcdeveloper-toolsvk973qbr2m4b7gxn4ey3q4eyg3n82tckclatestvk971p6y022gay5y715wn01qwkh82w2d8openclawvk973qbr2m4b7gxn4ey3q4eyg3n82tckc
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
🤖 Clawdis
Binsnode, npm