daily-paper-generator
Pass
Audited by Gen Agent Trust Hub on May 17, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and interacts with several well-known academic services, such as arxiv.org, nber.org, and major academic journal publishers. These interactions are necessary for the skill's discovery and retrieval functions.
- [PROMPT_INJECTION]: The skill processes paper abstracts and metadata from external sources to generate summaries, presenting a surface for indirect prompt injection. 1. Ingestion points: External research abstracts are fetched via web navigation and the scripts/arxiv_search.py script. 2. Boundary markers: The instructions do not specify explicit delimiters to prevent instructions embedded within paper abstracts from influencing the agent's behavior. 3. Capability inventory: The agent is authorized to write files to the local filesystem and perform web navigation. 4. Sanitization: No explicit sanitization of the external abstract text is performed before it is processed by the model.
- [COMMAND_EXECUTION]: The skill includes a helper script, scripts/arxiv_search.py, which queries the arXiv API using the feedparser library. The script constructs queries using standard methods. A discrepancy exists between the script's default search keywords (EEG) and the skill's primary focus (Economics), which appears to be a documentation inconsistency.
Audit Metadata