yichen-wechat-windows-reader

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The test file tests/test_snapshot_reader.py uses subprocess.run to execute the CLI tool during automated testing. This is a standard development practice using fixed, local paths.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from user-supplied SQLite databases. It mitigates the risk of instructions embedded in chat content being executed by the AI by wrapping all outputs in untrusted_snapshot_data JSON markers and markdown code fences, accompanied by explicit instructions for the agent to treat the data as untrusted.
  • [EXTERNAL_DOWNLOADS]: The project requires the zstandard Python package for decompressing database fields. The dependency is securely managed through requirements.lock with mandatory hash verification and binary-only installation constraints, fetching from the official PyPI registry.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 09:07 AM
Security Audit — agent-trust-hub — yichen-wechat-windows-reader