soia-pkm-interpret-article-analysis

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides installation instructions that reference the author's GitHub repositories ('soia-team') via the platform's plugin marketplace and the npx package runner. These are documented steps for fetching components within the same vendor ecosystem.\n- [COMMAND_EXECUTION]: The agent is instructed to read local Markdown files and write new interpretation files within a specific user vault directory. These operations are essential to the skill's primary function and are performed without elevated privileges.\n- [PROMPT_INJECTION]: As the skill ingests external article data, it possesses an indirect prompt injection surface; however, the skill design minimizes risk through clear boundaries.\n
  • Ingestion points: The skill reads article content from files in the vault as defined in SKILL.md.\n
  • Boundary markers: The skill maintains a file-level separation, generating output in a new file with a specific suffix (-AI解读.md) instead of appending to the source.\n
  • Capability inventory: The skill uses basic file system read and write capabilities.\n
  • Sanitization: The instructions include explicit constraints ('绝不碰原文') that prevent the agent from modifying or overwriting the original document or the user's private opinion sections.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 08:08 AM
Security Audit — agent-trust-hub — soia-pkm-interpret-article-analysis