NYC

peer-review-initiator

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION] (LOW): The code-reviewer.md template interpolates external data such as {DESCRIPTION} directly into agent instructions without delimiters. 1. Ingestion points: code-reviewer.md placeholders ({WHAT_WAS_IMPLEMENTED}, {PLAN_OR_REQUIREMENTS}, {DESCRIPTION}, {PLAN_REFERENCE}). 2. Boundary markers: Absent. No delimiters distinguish instructions from user-provided content. 3. Capability inventory: Shell command execution via git diff. 4. Sanitization: Absent. No escaping or validation is performed on inputs.
  • [COMMAND_EXECUTION] (LOW): The skill uses shell commands with variable interpolation (e.g., git diff {BASE_SHA}..{HEAD_SHA}). While intended for git SHAs, these could be vectors for shell injection if inputs are not strictly validated as hex strings.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:01 PM