skills/duong/dotfiles/review/Gen Agent Trust Hub

review

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION] (SAFE): The skill utilizes shell commands (git rev-parse, find) to discover project-specific rule files. These commands are limited to local path discovery and do not involve remote execution or system modification.
  • [PROMPT_INJECTION] (LOW): The skill is vulnerable to Indirect Prompt Injection (Category 8). It instructs the agent to discover and read the contents of various files (AGENTS.md, .cursorrules, etc.) to guide its review process. An attacker could place malicious instructions in these files to hijack the agent's logic during the review session.
  • Ingestion points: File contents of AGENTS.md, AGENT.md, .cursor/rules/*.mdc, and .cursorrules.
  • Boundary markers: Absent. The instructions do not specify delimiters or warnings to ignore embedded instructions in the rule files.
  • Capability inventory: File reading, local path discovery via shell, and reasoning/reporting.
  • Sanitization: None. The agent is directed to treat the discovered files as authoritative project rules.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:29 PM