Skillv1.0.0

ClawScan security

Omni-Agent Builder (OpenClaw) · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignMar 5, 2026, 12:47 AM

Verdict: Benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's requested resources, files, and runtime instructions align with its stated purpose of scaffolding and validating OpenClaw agent workspaces; there are no surprising network calls, credential requests, or installers.
Guidance: This skill appears coherent and focused on generating safe OpenClaw workspaces. Before using it: (1) inspect the generated files locally and remove any secrets before committing or sharing; (2) run scripts (scaffold + validate) in a safe directory — they create and write files in the target directory; (3) if you enable the agent to run local CLI commands (openclaw or others), ensure you don't grant it tokens/credentials or channel integrations (Telegram/Slack/etc.) without explicit, manual configuration and review; (4) use the included validate-workspace.py to confirm guardrails are present. If you have sensitive data or private keys, do not put them in the workspace or request the agent to persist them into MEMORY.md.

Review Dimensions

Purpose & Capability: okName/description (workspace scaffolding, templates, guardrails) match the provided files and scripts. The two scripts (scaffold + validate) and the references support the stated purpose; no unrelated credentials, binaries, or install steps are requested.
Instruction Scope: okSKILL.md instructs only workspace-related actions: asking the user clarifying questions, generating workspace files, and including safety guardrails. It does ask about channels and whether the agent should run local openclaw CLI commands, but it does not instruct reading unrelated system files, extracting secrets, or contacting third-party endpoints. It also instructs printing file contents (for deliverables) — users should avoid including secrets in inputs to this generation step.
Install Mechanism: okNo install spec is present (instruction-only skill), and included scripts are small, local, and benign. Nothing is downloaded from external URLs or written to nonstandard system paths by the skill itself.
Credentials: okThe skill declares no required environment variables, no primary credential, and no config paths. The scripts and SKILL.md explicitly advise not to store credentials in the workspace. This is proportionate to the scaffolding/validation task.
Persistence & Privilege: okalways is false and the skill is user-invocable. The skill does create workspace files when its scaffold script is run, which is expected; it does not request permanent platform privileges or modify other skills or global agent configuration.