review-diff

Pass

Audited by Gen Agent Trust Hub on Nov 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection because it ingests and processes untrusted data (git diffs and source code) and possesses file-writing capabilities.
  • Ingestion points: The skill reads the output of git diff and git status, as well as the contents of files pointed to by the diff, as described in SKILL.md and reviewer.md.
  • Boundary markers: While the skill uses a structured brief to instruct subagents, it lacks robust delimiters or explicit instructions to the subagents to ignore natural language instructions that might be embedded in code comments or string literals within the diff.
  • Capability inventory: The skill has the authority to read arbitrary files in the repository and the capability to apply 'fix-now' or 'fix-trivial' changes directly to the filesystem in Step 6 of SKILL.md.
  • Sanitization: No sanitization or filtering is performed on the diff content before it is interpolated into the prompt for the reviewer subagents.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 4, 2026, 03:10 AM