Skillv1.0.0

ClawScan security

查询高驰（COROS）运动手表的跑步运动数据 · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

SuspiciousMar 10, 2026, 3:23 AM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The code and instructions match the stated purpose (querying COROS running activities) but the skill fails to declare required credentials in its metadata and asks you to store a password hash in a .env file — this mismatch and presence of credential-handling warrants caution.
Guidance: This skill's code appears to do what it claims (login with COROS credentials and fetch activities), but the package metadata fails to declare the required environment variables (COROS_ACCOUNT, COROS_PASSWORD). Before installing or running it: 1) Verify the COROS endpoints (teamcnapi.coros.com) are legitimate for your region; 2) Do not commit a filled .env to any repository — treat it as a secret; 3) Prefer to use an API token or app-specific credential if COROS supports one rather than your main account password; 4) If you must use a password hash, consider creating a throwaway account or rotating the password after testing; 5) Run the code in an isolated environment (container) and inspect network traffic if possible; 6) Ask the author to update the skill metadata to explicitly list required env vars and explain why an MD5 of the password is required. These steps will reduce the risk of accidental credential exposure.

Review Dimensions

Purpose & Capability: okThe name/description describe querying COROS activity data and the included code (coros.js, util.js) implements logging in and fetching activity lists from COROS endpoints (teamcnapi.coros.com). The requested dependencies (axios, dotenv) and utility functions (MD5 hashing, distance summation) are appropriate for the stated purpose.
Instruction Scope: concernSKILL.md and coros.js instruct the agent to read credentials from scripts/.env (COROS_ACCOUNT and COROS_PASSWORD) and to call COROS APIs; these runtime instructions are narrowly scoped to the described task. However the skill explicitly instructs storing an MD5-hashed password in .env and the package includes a .env template — reading/writing credential files is sensitive and the metadata does not declare these env requirements (mismatch noted).
Install Mechanism: okThere is no install spec (instruction-only) but source files and package.json are provided. Dependencies are standard (axios, dotenv) and package-lock references common npm packages (mirrors.tencent.com entries appear to be a registry mirror). No arbitrary download URLs or archive extracts are present.
Credentials: concernThe skill actually requires COROS_ACCOUNT and COROS_PASSWORD (MD5) environment variables and includes a scripts/.env file, yet the registry metadata states 'Required env vars: none'. Requiring an account credential is reasonable for this API integration, but the omission from metadata is an inconsistency and the skill asks the user to persistively store a hashed password — both are privacy-sensitive and should be explicitly declared and justified.
Persistence & Privilege: okSkill flags are default (always: false, user-invocable: true). It does not request system-wide config changes, no always:true. It does read a local .env in its own folder (normal for a script), and does not attempt to modify other skills or global agent settings.