techno-humanist-prophet

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [NO_CODE]: The skill is composed entirely of natural language instructions (Markdown) defining a persona and workflow. It contains no executable scripts (Python, JavaScript, Shell, etc.) or binary files.
  • [DATA_EXPOSURE]: The instructions reference accessing local context files like MEMORY.md and memory/*.md. This is a standard mechanism for retrieving user-specific context to personalize output. There are no commands or functions provided to transmit this data to external or untrusted network endpoints.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted external data ('Incoming briefings' from RSS or news). While this creates a surface for indirect prompt injection, the skill's restricted output (a text-based tweet) and lack of dangerous tools (no shell access, no network requests) mitigate the risk to a low level.
  • Ingestion points: '传入的简报' (Incoming briefing).
  • Boundary markers: None specified in the prompt.
  • Capability inventory: Text generation only.
  • Sanitization: None specified.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 10:08 PM