Skillv0.1.1

ClawScan security

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

Scanner verdict

BenignMar 6, 2026, 12:55 PM

Verdict: benign
Confidence: medium
Model: gpt-5-mini
Summary: The skill's code and instructions are coherent with its stated purpose (continuous network diagnostics against Telegram Bot API); nothing in the package demands unrelated credentials or privileged system changes, but you should review config/log handling and run it isolated until you confirm behavior.
Guidance: This package appears to do what it says, but take these precautions before deploying: (1) Review the full scripts/netdiag.py source yourself (or have someone you trust) to confirm there are no unexpected network endpoints or hidden behaviors. (2) Keep your Telegram bot token and personal chat id out of world-readable locations; prefer a restricted config file and OS-level secret storage. (3) Ensure logging.redact_sensitive_fields is true before running; if you must share logs externally (AI analysis, etc.), mask tokens and PII first. (4) Run the worker as an external background process (the SKILL.md recommendation) rather than in-process to avoid impacting OpenClaw runtime. (5) Start in a sandboxed host or staging environment to observe actual network traffic and any subprocess invocation (ping/traceroute/dig) before deploying to production. (6) If you supply a proxy_url or NETDIAG_CONFIG via environment, double-check those values to prevent accidental routing of telemetry to an unintended server.

Review Dimensions

Purpose & Capability: okName/description match the files and runtime instructions: the worker probes DNS, ICMP, traceroute, MTU, and exercises the Telegram Bot API for delivery verification. Required artifacts (bot token, personal chat id) are consistent with the stated functionality and no unrelated cloud credentials or system-wide accesses are requested.
Instruction Scope: noteSKILL.md stays focused on diagnostics and gives explicit CLI commands and background/foreground modes. It recommends both external-process and in-process integration; the latter is explicitly flagged as higher-risk. Attention: the workflow and references encourage later AI-based log analysis and may involve sharing log slices externally — ensure tokens/PII are redacted before exporting logs.
Install Mechanism: okNo install spec is provided (instruction-only), so nothing is fetched or executed at install time. The runtime requires Python 3.11+ and common networking tools (dig/ping/traceroute) which are proportional to the task.
Credentials: noteThe skill does not ask for unrelated environment credentials in registry metadata. It expects a config file containing a Telegram bot token and personal chat id (or NETDIAG_CONFIG can point to a config), which is appropriate. Be aware the tool writes detailed JSON logs that could include sensitive headers or tokens if redaction is disabled; default config sets redaction=true, but you should verify it's enabled.
Persistence & Privilege: okThe skill is not always-enabled and is user-invocable. It runs as a long-running process (writes PID file, logs) which is normal for a monitoring worker. The in-process integration option is documented and acknowledged to be riskier; prefer the external-process option to maintain isolation.