antigravity-lazy-packs

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.70). The links point to a mix of a known site (cc.lifehacker.tw) and a set of GitHub repos from an individual/unknown account (mathruffian-dot) plus an external vendor (firecrawl.dev); while there are no direct .exe downloads, the skill docs instruct using npx/npm/pip (commands that can execute arbitrary third‑party code) and the GitHub account is not a well‑known/verified vendor, so this represents a moderate supply‑chain risk and should be treated with caution.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The GitHub repo https://github.com/mathruffian-dot/antigravity-lazy-pack is explicitly fetched at runtime via commands like "npx skills add mathruffian-dot/antigravity-lazy-pack" (and the README tells the agent to read that repo), which installs remote skill files (SKILL.md) that directly define/modify agent prompts/behavior and may run remote code, so it is a runtime external dependency that controls the agent.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.90). It directs the agent to run global npm/npx installs (npx ... -g -y) and to "execute" instructions from SKILL.md (arbitrary commands), which modifies the host system and may require elevated privileges or run unsafe code, so it poses a high risk of compromising machine state.

MEDIUM W021: Hidden or invisible Unicode characters detected (potential obfuscation or prompt injection).

• Hidden Unicode characters detected (1 type(s) found)