ccc-ultracode

Pass

Audited by Gen Agent Trust Hub on Jun 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses Bash and a Workflow tool to automate tasks such as codebase migrations and repo-wide audits.\n- [REMOTE_CODE_EXECUTION]: Orchestrates complex tasks by executing specific JavaScript workflow files (e.g., ccc-migrate.workflow.js) stored in the plugin's root directory (${CLAUDE_PLUGIN_ROOT}).\n- [PROMPT_INJECTION]: The skill processes external repository data (files and diffs), presenting a surface for indirect prompt injection. This is an inherent property of auditing tools that read untrusted code.\n
  • Ingestion points: Processes repository files and branch diffs via the Read tool.\n
  • Boundary markers: Does not specify delimiters to separate analyzed code from instructions.\n
  • Capability inventory: Can execute shell commands via Bash and launch sub-agents via the Workflow tool.\n
  • Sanitization: Content analysis is performed directly on the source files without explicit pre-processing or sanitization filters.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 20, 2026, 07:21 PM
Security Audit — agent-trust-hub — ccc-ultracode