cali-product-codebase-critique

Pass

Audited by Gen Agent Trust Hub on Jul 3, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses shell commands like find, git diff, and wc to analyze the codebase structure and determine the audit depth based on the number of changed files.
  • [EXTERNAL_DOWNLOADS]: The reference documentation points to various external plugins and NPM packages needed for full functionality, including tools from well-known services like Cursor (github.com/cursor/plugins) and other third-party repositories (PriNova/pi-agent-codebase-workflows, @plannotator/*, @juicesharp/*).
  • [PROMPT_INJECTION]: The skill processes untrusted source code from a user-provided directory ({INPUT_PATH}), creating a surface for indirect prompt injection.
  • Ingestion points: Codebase files are read and processed by the reviewer subagent as described in SKILL.md.
  • Boundary markers: The subagent task description does not specify explicit delimiters or 'ignore' instructions for the analyzed code.
  • Capability inventory: The skill uses subagent for delegation, find for file discovery, and write for saving the final report.
  • Sanitization: There are no mentioned mechanisms for sanitizing or escaping instructions found within the analyzed source code.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 3, 2026, 12:53 PM
Security Audit — agent-trust-hub — cali-product-codebase-critique