novelty-assessment

Pass

Audited by Gen Agent Trust Hub on Apr 21, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes local Python scripts (novelty_check.py, search_semantic_scholar.py) located in the ~/.claude/skills/ directory. These scripts are used for processing research ideas and querying academic APIs, which aligns with the skill's stated purpose.
  • [EXTERNAL_DOWNLOADS]: The skill performs network operations to retrieve data from well-known scholarly services including Semantic Scholar, arXiv, and OpenAlex. These are trusted sources for academic research data.
  • [PROMPT_INJECTION]: The skill utilizes persona-based prompting (e.g., 'Harsh Critic') and structured protocols for multi-round assessment. These instructions are task-specific and do not attempt to bypass safety filters, extract system prompts, or override agent constraints.
  • [CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or credentials were found. The skill references external APIs but does not expose sensitive authentication data.
  • [DATA_EXFILTRATION]: There is no evidence of data exfiltration. The search queries generated are based on the research idea and are sent only to legitimate scholarly databases.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 21, 2026, 07:28 AM