Skillv1.0.0

ClawScan security

Legal case wdy · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

SuspiciousApr 29, 2026, 9:28 AM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The skill's behavior (sending user case text to a third‑party API requiring WENDAOYUN_API_KEY) matches its purpose, but the package metadata omits the required credential and there are clear privacy risks from transmitting case details to an external service.
Guidance: Before installing or enabling this skill, consider the following: - The skill will send the text you provide (possibly including names, amounts, injuries, or other sensitive details) to a third‑party API (h5.wintaocloud.com). Only use it with explicit consent and avoid sending full PII if you want to protect privacy — redact or anonymize when possible. - The SKILL.md requires an environment variable WENDAOYUN_API_KEY, but the registry metadata does not declare it. Confirm with the publisher (or the skill manifest) before trusting that the platform will prompt for or protect this key. - Verify the legitimacy and privacy/security practices of the external service (wintaocloud/open.wintaocloud). Confirm terms of use, retention policy, and whether submitted data is stored or used for model training. - If you must test, try non-sensitive/dummy queries first and rotate/revoke the API key if you suspect misuse. - Prefer skills with clear provenance (homepage, owner contact) and explicit declared env requirements. If the publisher cannot justify the metadata mismatch or the third‑party provider, treat the skill cautiously or decline to install.

Review Dimensions

Purpose & Capability: noteThe skill claims to perform legal retrieval and similar-case lookup and its SKILL.md calls an external legal/case API — this is coherent with the stated purpose. However the registry metadata does not declare the required WENDAOYUN_API_KEY environment variable that the runtime instructions insist on, creating an inconsistency.
Instruction Scope: concernRuntime instructions direct the agent to send users' case descriptions (which may include names, amounts, injuries, etc.) to a third‑party endpoint (https://h5.wintaocloud.com) via curl. The SKILL.md explicitly warns about privacy, but transmitting PII to an external service is an inherent privacy/data‑leak risk and should be disclosed to and consented by users. Instructions otherwise stay within the task (no unrelated file reads or other secrets accessed).
Install Mechanism: okInstruction‑only skill with no install spec or code files — lowest install risk. No downloads or disk writes are described.
Credentials: concernThe SKILL.md requires a single API key (WENDAOYUN_API_KEY), which is proportionate for calling the described third‑party API. However the registry metadata lists no required env vars — this mismatch is an incoherence and could lead to unexpected behavior or hidden requirements. The skill also instructs not to echo the key (good), but the missing declaration in metadata is a red flag.
Persistence & Privilege: okNo always:true, no installs, no changes to other skills or system config — the skill does not request elevated persistence or privileges.