Security audit
Security checks for vulnerabilities and agentic risk
The skill’s policy comparison workflow is clearly scoped, with disclosed browser setup and optional draft filling that requires user authorization.
Use a dedicated browser profile and explicitly select its browserId. Leave draft filling disabled unless wanted. Review installer and dependency versions before setup; the supplied skill files do not establish the safety of downloaded browser software.
The only supplied operation installs a skill; it does not perform the declared policy comparison. Installation is a materially different, state-changing action outside the described read-only workflow. The installed skill's subsequent behavior cannot be determined from this chunk.
The command performs a skill-installation operation, which changes the local skill setup and is materially different from reviewing public policy pages. Although the repository name is related to shipping-policy comparison, that does not establish that the declared workflow is implemented. This assessment is limited to the supplied command; the behavior of any downloaded skill is not shown.
The command npx skills add xxjrq/shop-shipping-policy-compare-free can download and execute an unpinned version of the skills npm package, exposing installation to compromised future releases with the invoking user's privileges. This is a supply-chain risk rather than an MCP-server configuration issue; it is a conventional installation command with no evidence of malicious intent, but the skill's read-only browsing scope does not constrain installer execution.
The installation command invokes npx skills without a version pin, allowing npm to resolve and execute a mutable CLI version with the invoking user's permissions. A compromised future release could therefore execute malicious code before the skill's read-only browser restrictions apply; however, this is ordinary installation guidance with no evidence of malicious intent, and the finding's characterization as an MCP server reference is inaccurate.
The manifest at L03 limits the skill to read-only comparison of public shipping or after-sales policy pages. However, L25 introduces optional draft filling, and L38 instructs the agent to fill drafts when authorized, which modifies page content rather than merely reading policy information. Requiring authorization and prohibiting submission limit the action but do not reconcile it with the manifest's read-only promise.
Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
No suspicious patterns detected.