zhihu-fetcher

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The script scripts/fetch_zhihu_batch.py uses subprocess.run to implement a retry mechanism. It invokes its own script file with specific arguments (--retry-failed) to resume interrupted batch tasks. The command is constructed using the script's absolute path and internal flags, which restricts execution to the skill's own logic and prevents arbitrary command injection.
  • [EXTERNAL_DOWNLOADS]: The skill downloads images from Zhihu domain and relies on Playwright to download necessary browser binaries (Chromium). These downloads are directed to well-known and expected services required for the scraping and archiving functionality.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external Zhihu articles (titles and content). It implements filename sanitization using regular expressions (e.g., re.sub(r'[\\/:*?"<>|]', '_', title)) to prevent path traversal or illegal character issues when creating local files. It also utilizes standard YAML frontmatter boundaries (---) to safely encapsulate metadata in the generated Markdown notes.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 05:04 PM
Security Audit — agent-trust-hub — zhihu-fetcher