best-practices-guidelines
Pass
Audited by Gen Agent Trust Hub on Mar 3, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill is designed to review, refactor, and improve user-provided code, which creates a surface for indirect prompt injection where malicious instructions embedded in the analyzed code could influence agent behavior.
- Ingestion points:
SKILL.md(Review code for guideline compliance). - Boundary markers: Absent. The instructions do not specify delimiters or guidelines to ignore instructions within the processed code.
- Capability inventory: The skill is granted
Read,Write, andEdittools as perSKILL.mdfrontmatter. - Sanitization: Absent. There is no validation or filtering of the code content before analysis.
- [COMMAND_EXECUTION]: The
SKILL.mdfile includes a mandatory 'Memory Protocol' section that instructs the agent to execute a shell command (cat .claude/context/memory/learnings.md) to retrieve state. - Evidence: File
SKILL.mdline 61. While the command is a read operation on a specific path, it represents an instruction for the agent to execute local system commands.
Audit Metadata