Skills
skills/oschina/gitee-agent-skills/review-pr/Gen Agent Trust Hub

review-pr

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted code and comments from external Pull Requests, which creates an attack surface for indirect prompt injection.\n
  • Ingestion points: Pull request details, diff files, file content, and existing comments are fetched from Gitee and ingested into the agent's context (SKILL.md, Steps 1-3).\n
  • Boundary markers: The instructions lack explicit delimiters or specific warnings to ignore instructions found within the processed PR data.\n
  • Capability inventory: The agent can post comments to Gitee via comment_pull and execute local git and file system commands (SKILL.md, Step 2 & 7).\n
  • Sanitization: No sanitization or validation of the PR data is performed before processing.\n- [COMMAND_EXECUTION]: The skill instructs the agent to execute local system commands.\n
  • Evidence: The agent is guided to use local git commands and file access to obtain broader context if a local repository clone is available (SKILL.md, Step 2).
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 02:37 PM