The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands to run a Python script (semantic_scholar_fetch.py) for paper searches, passing user-defined queries and parameters as arguments. This is the primary method for retrieving data.
[EXTERNAL_DOWNLOADS]: The skill connects to the Semantic Scholar API (api.semanticscholar.org), which is a well-known service for academic literature, to perform searches and fetch metadata.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it handles and displays external content (paper abstracts, titles, and TLDRs) which could contain malicious instructions. * Ingestion points: Data from the Semantic Scholar API enters the agent context during paper searches. * Boundary markers: The skill instructions specify using Markdown tables and headers to organize and separate external content. * Capability inventory: The agent is equipped with Bash(*), Read, and Write tools, allowing for command execution and file operations. * Sanitization: The instructions do not define any sanitization or filtering processes for the data retrieved from the external API.

semantic-scholar