code-review-assistant

Pass

Audited by Gen Agent Trust Hub on Oct 8, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process git diffs and pull request content, which are untrusted external data sources. This creates a vulnerability where an attacker could include hidden instructions within a code change to manipulate the agent's review output or behavior.
  • Ingestion points: The diff input parameter in skill.yaml serves as the primary entry point for untrusted data.
  • Boundary markers: No explicit delimiters or instructions to ignore embedded commands are defined in the skill configuration.
  • Capability inventory: The agent performs static analysis and generates review comments based on the provided diff.
  • Sanitization: There is no evidence of input validation, escaping, or filtering to mitigate the risk of instructions embedded within the code diffs.
  • [NO_CODE]: The skill only contains markdown documentation and a YAML configuration file. It does not include any Python scripts, JavaScript files, shell scripts, or other executable binaries, which significantly limits the risk of direct malicious actions like credential theft or remote execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 8, 2026, 10:24 AM