security-scan

Warn

Audited by Gen Agent Trust Hub on Mar 24, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to download and install the ecc-agentshield package globally via npm or execute it directly using npx from the public npm registry.
  • [REMOTE_CODE_EXECUTION]: Execution of the ecc-agentshield tool via npx constitutes running code fetched from a remote source at runtime.
  • [COMMAND_EXECUTION]: The skill executes multiple shell commands to perform scans, including --fix operations that modify local configuration files and an --opus mode that interacts with external AI models.
  • [PROMPT_INJECTION]: The skill exposes an indirect prompt injection surface by ingesting and processing untrusted data from user-controlled files like CLAUDE.md, .claude/settings.json, and agent definitions.
  • Ingestion points: Files within the .claude/ directory and project root.
  • Boundary markers: None identified in the provided instructions.
  • Capability inventory: Command execution (npx), file modification (--fix), and network access (implied by API key requirement for Opus mode).
  • Sanitization: None described in the skill metadata or instructions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 24, 2026, 07:42 AM
Security Audit — agent-trust-hub — security-scan