github.com/Sanskar015/claude-skills-supercharged · main · scanned 2 h ago · 6 config files
7 agent components in Sanskar015/claude-skills-supercharged: an installer lure (malware delivery pattern), 4 unvetted.
Components70 MCP servers
High blast radius1credentials, shell or unvetted code
Credentials handed to agents0none seen
Not vetted4of 4 local servers, skills and plugins
| Component | Where | Verdict | Blast radius |
|---|---|---|---|
Installer lure (malware delivery pattern)Installer lure · any agent | README.md | malware pattern | high
|
api-securitySkill · Claude Code | .claude/skills/api-security/SKILL.md | not vetted | low
|
git-workflowSkill · Claude Code | .claude/skills/git-workflow/SKILL.md | not vetted | low
|
python-best-practicesSkill · Claude Code | .claude/skills/python-best-practices/SKILL.md | not vetted | low
|
skill-developerSkill · Claude Code | .claude/skills/skill-developer/SKILL.md | not vetted | low
|
slash command: wrapInstructions · Claude Code | .claude/commands/wrap.md | clean | low
|
UserPromptSubmitHook · Claude Codeskill-activation-prompt.sh | .claude/settings.json | local command | low
|
What to do about it
- Do not download or run anything from this repository. Its README sends readers to a committed installer and tells them to click past Windows' own warning; that is how the Agent Lure Stealer campaign (Nov 2026) delivers an infostealer. Real skills and MCP servers install from source, never from a ZIP.
- Unvetted means we have no sandbox run for that exact package. Ask your agent to check before installing: claude mcp add killerskills -- npx -y vet-skills mcp
- For a whole organization (every repo and laptop, rescanned daily, with removal across machines), see Discovery for teams.
Share: https://killerskills.ai/scan/github/Sanskar015/claude-skills-superchargedRescan nowScan another