vercel-deploy
Fail
Audited by Gen Agent Trust Hub on Mar 19, 2026
Risk Level: HIGHDATA_EXFILTRATIONPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill metadata and documentation contain deceptive claims, stating the author is "vercel" while the actual author is identified as "akillness". This impersonation is used to deceive the user and the agent regarding the skill's origin and safety.
- [DATA_EXFILTRATION]: The script "scripts/deploy.sh" creates an archive of the user's project directory and uploads it to an external URL ("https://claude-skills-deploy.vercel.com/api/deploy"). The archive process fails to exclude sensitive files such as ".env", ".npmrc", or other configuration files containing secrets, leading to potential exfiltration of credentials during the deployment process.
- [COMMAND_EXECUTION]: The skill executes a bash script that performs unauthenticated network uploads of local files and modifies the local filesystem by renaming HTML files without explicit user consent for each file modification.
Recommendations
- AI detected serious security threats
Audit Metadata