security-scan

Warn

Audited by Gen Agent Trust Hub on Oct 14, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill directs the agent to download and install the ecc-agentshield package from the npm registry and execute it using npx. This package is hosted in a third-party GitHub repository (affaan-m/agentshield) which is not associated with a known trusted organization.
  • [COMMAND_EXECUTION]: Instructions involve executing various shell commands to perform project-wide scans, initialize configuration files, and apply automated fixes to existing security settings.
  • [INDIRECT_PROMPT_INJECTION]: The skill audits files such as CLAUDE.md, project hooks, and agent definitions that may be populated with content from untrusted external sources. These ingestion points lack explicit boundary markers or sanitization instructions, creating a surface where malicious instructions embedded in project files could influence the agent's behavior during the audit.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Oct 14, 2026, 02:59 PM