The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands using the pyzotero CLI tool and the jq utility to perform library searches, retrieve item metadata, and process JSON results (SKILL.md, references/jq-recipes.md).
[EXTERNAL_DOWNLOADS]: The instructions prompt the user to install the pyzotero package from PyPI, which is the official library maintained by the skill's author (SKILL.md, README.md).
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes untrusted research data (SKILL.md, references/research-patterns.md).
Ingestion points: Retrieves paper abstracts and indexed full-text content from the Zotero library and Semantic Scholar.
Boundary markers: Absent; the skill does not specify delimiters for separating paper content from instructions during analysis.
Capability inventory: Can execute CLI commands via pyzotero and jq to interact with local and remote citation data.
Sanitization: No explicit sanitization or filtering of research content is performed before the agent analyzes it.

zotero-search