skills/buiducnhat/co-drew/cd-review/Gen Agent Trust Hub

cd-review

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes shell commands like git diff and file reading to inspect the workspace. These operations are essential for the tool's primary purpose of providing feedback on code changes and are conducted within the local context of the project.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it analyzes untrusted data (uncommitted code and diffs) that may contain adversarial instructions meant to manipulate the agent's review report.\n
  • Ingestion points: SKILL.md (Step 1: Gather Context reads file contents and git diff output)\n
  • Boundary markers: Absent (the prompt does not use delimiters to isolate untrusted data from instructions)\n
  • Capability inventory: Shell access (git) and file reading access\n
  • Sanitization: Absent (the skill does not filter the content of the files before analysis)
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 11:23 AM
Security Audit — agent-trust-hub — cd-review