Skillv0.1.0

ClawScan security

Allstar Link node control ASL3 (ASL3 Node Control) · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignFeb 11, 2026, 8:58 AM

Verdict: benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's code, instructions, and required environment variables are coherent with its stated purpose (a client for an AllStar Link ASL3 agent); nothing obvious is requesting unrelated credentials or installing arbitrary remote code.
Guidance: This skill appears to be what it claims: a client for a self-hosted ASL agent. Before installing or running it, consider these practical checks: (1) Verify the ASL agent you will contact is your own Pi (ASL_PI_IP) and that ASL_API_KEY is scoped/rotated appropriately — do not reuse broad keys. (2) The shell script has a hardcoded default IP (100.116.156.98); always set ASL_PI_IP/ASL_API_BASE explicitly so you don't accidentally target another host. (3) The Python client requires the 'requests' package; install it in your environment (pip install requests) or run in an environment that provides it. (4) The client creates per-user files under ~/.openclaw/state/asl-control and will read the secrets file you source (e.g. ~/.config/secrets/api-keys.env) — keep that file protected. (5) There is no homepage or published source provenance; if you do not trust the unknown author, review the included scripts (which are short and readable) line-by-line before running. (6) Check the ASL agent audit log on the Pi (/opt/asl-agent/audit.log) after using the client to confirm only expected actions occurred.

Review Dimensions

Purpose & Capability: okName/description say 'control AllStar Link nodes via REST API' and the included shell/Python clients do exactly that: they call a local/remote ASL Agent HTTP API with an X-API-Key. The required binaries (python3) and env vars (ASL_PI_IP, ASL_API_KEY) match the purpose. A minor note: the shell wrapper has a hardcoded ASL_PI_IP default (100.116.156.98), but the SKILL.md explicitly instructs you to set ASL_PI_IP and ASL_API_KEY.
Instruction Scope: okSKILL.md instructs the agent to source a secrets file and run the provided Python script. The scripts only access the declared env vars (plus optional ASL_API_BASE, ASL_STATE_DIR) and per-user state files under ~/.openclaw/state/asl-control. They do not attempt to read unrelated system credentials or contact external endpoints other than the ASL agent base URL.
Install Mechanism: okNo install spec; this is instruction-plus-scripts only. Nothing is downloaded or extracted from external URLs by the skill itself.
Credentials: okOnly ASL_PI_IP and ASL_API_KEY (plus optional ASL_API_BASE and ASL_STATE_DIR) are required — appropriate for a REST API client. No unrelated tokens or broad credentials are requested. One operational omission: the Python client depends on the 'requests' package but the skill metadata only requires 'python3' (this is usability/packaging friction, not a secret mismatch).
Persistence & Privilege: okalways is false and the skill writes only per-user state under ~/.openclaw/state/asl-control. It does not request system-wide changes or modify other skills. Model-autonomous invocation is allowed by default (not a problem here) — nothing else in the package amplifies that privilege.