The Agent Skills Directory

[COMMAND_EXECUTION]:
The skill utilizes curl to interact with external academic APIs for searching and retrieving paper metadata.
The skill uses uvx to execute the arxiv2markdown command-line tool, which converts ArXiv content into markdown format.
[EXTERNAL_DOWNLOADS]:
The skill fetches data from several external domains including api.alphaxiv.org, api.openalex.org, arxiv.org, and ar5iv.org.
It dynamically installs and runs the arxiv2markdown package from a public registry via the uvx tool.
[PROMPT_INJECTION]:
The skill is susceptible to Indirect Prompt Injection because it ingests untrusted external content into the agent's context.
Ingestion points: Full-text content from ArXiv papers, alphaXiv overviews, and abstracts from OpenAlex are loaded into the agent's active session (e.g., via paper.md).
Boundary markers: The instructions do not implement any boundary markers or special delimiters to warn the agent that the fetched content should be treated as untrusted data.
Capability inventory: The skill environment allows for network requests (curl), running external software (uvx), and writing to the local file system.
Sanitization: No sanitization or filtering logic is present to identify or neutralize potential instructions hidden within the research papers.

find-and-read-papers