skills/ascend-ai-coding/awesome-ascend-skills/external-cannbot-ops-ascendc-code-review/Gen Agent Trust Hub
external-cannbot-ops-ascendc-code-review
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data as core parameters defined in SKILL.md, specifically the code snippets and review rule descriptions. 1. Ingestion points: Parameters in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: Git operations (clone, fetch) via auxiliary Python scripts. 4. Sanitization: Absent. The lack of delimiters or sanitization steps means that instructions embedded within the analyzed code could potentially influence the agent's logic during the review process.
- [COMMAND_EXECUTION]: The skill provides Python scripts (scripts/clone_pr_source.py and scripts/get_gitcode_pr_diff.py) that execute git commands to fetch source code for analysis. These scripts use the subprocess module and pass arguments as lists, avoiding shell execution (shell=True). This implementation prevents standard command injection attacks and follows security best practices for system interactions.
- [EXTERNAL_DOWNLOADS]: The skill performs network operations to download source code and pull request data from gitcode.com. These operations are restricted to the GitCode domain through hardcoded whitelists in the scripts. As GitCode is a well-known developer platform relevant to the Ascend ecosystem and the skill's origin, these downloads are considered intended functionality and are documented neutrally.
Audit Metadata