aigc-down-skill

Pass

Audited by Gen Agent Trust Hub on Apr 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill's primary function is to provide instructions for academic text rewriting and style adjustment. It does not contain any executable code, shell commands, or network requests that would pose a security risk to the user's environment.
  • [DATA_EXPOSURE]: The skill uses a local directory session-memory/ to store session-specific metadata and logical summaries. This is used to maintain context during long-form rewriting tasks and does not involve the exfiltration of sensitive user credentials or system files.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (AIGC detection reports and user-provided paper content), which is an inherent attack surface for indirect prompt injection.
  • Ingestion points: User-provided HTML reports and academic text or LaTeX fragments as specified in SKILL.md and README.md.
  • Boundary markers: The skill does not implement specific delimiters or instructions to ignore potential commands embedded within the processed text.
  • Capability inventory: The agent's capabilities in this context are limited to text rewriting and writing session logs to the session-memory/ directory. No dangerous system-level or network capabilities are invoked.
  • Sanitization: There are no explicit input validation or sanitization routines defined for the external data processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 24, 2026, 01:58 PM
Security Audit — agent-trust-hub — aigc-down-skill