academic-writer
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill interacts with well-known academic platforms such as arXiv, Semantic Scholar, and DBLP through their public APIs to retrieve paper metadata and abstracts. These downloads are performed to facilitate the literature review process.
- [COMMAND_EXECUTION]: Local shell commands including
curlare used to fetch data from academic APIs. The fetched data is processed using Python one-liners to extract specific fields like titles and years, which is a routine practice for data transformation. - [INDIRECT_PROMPT_INJECTION]: By fetching and processing paper abstracts from external sources, the skill possesses an attack surface for indirect prompt injection. This is an inherent risk of the literature search functionality, where the agent processes third-party academic content.
- [DYNAMIC_EXECUTION]: The skill uses
python3 -cto execute short, static Python scripts for parsing XML and JSON data returned from search queries. These scripts are predefined within the skill instructions and are used for simple data manipulation rather than complex runtime code generation.
Audit Metadata