MAL-2026-18364: Malicious code in @zaip-ai/cli (npm)

--- -= Per source details. Do not edit below this line.=-

Source: amazon-inspector (0e44ea3cafd9835a85d3305b6aa9ed2f5400546938c6430c19a30cbdc5398936) @zaip-ai/cli@7.2.0 is a CLI package whose shipped files trigger keyword co-occurrence patterns (curl/wget/ping tokens, Buffer.from base64 decoding) without evidence of installer-side harm. The curl references in dist/services/codex-pre-tool-hook-template.test.js and dist/services/codex-pre-tool-policy-template.js appear inside test fixtures and policy templates that enumerate example command strings (including explicit placeholder hosts such as evil.test and example.com) used by a pre-tool hook/policy mechanism — the strings are policy data, not executed commands against those hosts. Buffer.from(..., 'base64').toString(...) usage in dist/services/global-codex-config.js and in a release-preparation script is ordinary base64 decoding of configuration/release data, with no eval, Function constructor, or dynamic code execution of the decoded bytes. The ping/POST/GET tokens in templates/.agents/tools/runtime/api-mcp-runtime.mjs come from an MCP runtime template describing HTTP method handling. No lifecycle scripts fetch remote code, no installer secret paths (~/.aws, ~/.ssh, ~/.npmrc, browser stores) are read, no hardcoded attacker C2 endpoint is bound to a network call, and no obfuscated destination is present.

Source: ghsa-malware (6871bd0117df69d113dc6e42ef94d417f866e24b30e2b2b0484a2cb934473443) Any computer that has this package installed or running should be considered fully compromised. All secrets and keys stored on that computer should be rotated immediately from a different computer. The package should be removed, but as full control of the computer may have been given to an outside entity, there is no guarantee that removing the package will remove all malicious software resulting from installing it.