AI Security Review
scanned 20h ago · by lpm-firewall-aiNo confirmed malicious attack surface is established. Risky primitives are part of a Hiver sandbox client and explicit examples, not automatic install or import behavior.
Decision evidence
public snapshot- examples/mcp-server/image/server.ts exposes MCP tools for bash/read/write/edit when that example server is explicitly run
- examples/mcp-server/index.ts spawns npx @modelcontextprotocol/inspector as an example command
- src/sandbox.ts provides user-invoked sandbox exec/writeFile/deleteFile API calls
- package.json has no preinstall/install/postinstall hooks
- src/controller.ts only contacts a configured Hiver gateway, defaulting to http://localhost:10000
- dist/index.js appears to be bundled src client code, not hidden install-time code
- process.env use is limited to HIVER_GATEWAY_URL and example-required API keys
- examples/utils/build.ts runs hiver bundle only in example workflows
- No credential harvesting, persistence, project config mutation, or exfiltration found
Source & flagged code
6 flagged · loading sourceThis package version adds a dangerous source file absent from the previous stored version; route for source-aware review.
dist/index.jsView on unpkgPackage source references shell execution.
examples/mcp-server/image/server.tsView on unpkg · L214A single source file combines environment access, network access, and code or shell execution; review context before blocking.
examples/mcp-server/image/server.tsView on unpkg · L1Package source invokes a package manager install command at runtime.
examples/mcp-server/index.tsView on unpkg · L3Package ships non-JavaScript build or shell helper files.
examples/local-filesystem-mount/skills/echo/echo.shView on unpkg