Skillv1.2.0

ClawScan security

Openclaw Genie · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignMar 7, 2026, 1:13 PM

Verdict: benign
Confidence: high
Model: gpt-5-mini
Summary: This is an instruction-only documentation skill about OpenClaw whose files and runtime instructions align with its stated purpose and do not request extra credentials or install code.
Guidance: This skill is a documentation/reference pack for OpenClaw and appears coherent with that purpose. It does contain example install commands (e.g., curl | bash) and references to local config paths and secret handling. Before installing or enabling it for autonomous use, consider: (1) only enable it if you want an agent that may recommend running local commands or reading OpenClaw config files; (2) if you are cautious about agents accessing local files or secrets, keep autonomous invocation disabled or restrict the agent's tool permissions; (3) verify any commands it suggests (especially one-liner installers) before running them on your machine.

Review Dimensions

Purpose & Capability: okName/description match the content: the skill is a large reference and quick-start for OpenClaw (installation, config, channels, memory, tools, deployment, etc.). It does not declare unrelated env vars, binaries, or install steps beyond documenting how to install OpenClaw itself, which is coherent with the skill's purpose.
Instruction Scope: noteSKILL.md is documentation and contains many example CLI commands, file paths (e.g., ~/.openclaw/*), and config snippets. The document does not itself instruct the agent to read or exfiltrate host secrets, but it does reference secrets and config locations (env files, token storage). Because skills are injected into prompts, an agent using this skill could legitimately suggest running commands or reading local files as part of troubleshooting — this is consistent with the purpose but worth noting for privacy/security.
Install Mechanism: okNo install spec and no code files to execute; the skill is instruction-only (Markdown). There is no third-party download or extraction step performed by the skill itself.
Credentials: okThe skill documents many OpenClaw-related environment variables and secret handling patterns (e.g., API keys, OPENCLAW_HOME), but the skill does not require or request any environment variables or credentials to be provided to the platform. The referenced secrets are appropriate for an OpenClaw guide.
Persistence & Privilege: okFlags are default: always:false, user-invocable:true, disable-model-invocation:false. There is no indication the skill will persist configuration or modify other skills; it's a read-only documentation skill and thus does not request elevated persistence privileges.