The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses uvx to download and execute the opencite tool from an external package registry. While uvx is a standard tool manager, it fetches code from the public PyPI repository at runtime.
[COMMAND_EXECUTION]: The skill provides instructions to execute several shell commands for searching literature, downloading papers, and generating bibliographies.
Evidence: uvx opencite search, uvx opencite batch-fetch, and a shell loop for uvx opencite lookup are used in SKILL.md.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core functionality of ingesting and processing untrusted external data.
Ingestion points: Research papers and metadata are fetched from external academic databases via opencite (referenced in SKILL.md).
Boundary markers: Absent. The instructions do not specify delimiters or provide warnings to the agent to ignore instructions embedded within the research papers.
Capability inventory: The skill possesses the ability to execute subprocesses via uvx and perform file system operations (writing review_refs.bib).
Sanitization: Absent. The skill lacks mechanisms to sanitize or validate the content of the retrieved papers before synthesis.

literature-review