The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it processes untrusted data from external sources.
Ingestion points: The skill ingests data from academic papers (via get_item_fulltext) and external search results (via WebSearch) to perform analysis and generate literature reviews.
Boundary markers: No explicit boundary markers or specific instructions to ignore instructions embedded in the analyzed papers are defined in the provided files.
Capability inventory: The skill has the capability to write local files (literature-review.md, research-proposal.md, references.bib), perform web searches, and manage items in a Zotero library through the Zotero MCP server.
Sanitization: There is no evidence of sanitization or filtering of the content retrieved from external sources before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill uses well-known academic services and repositories (arXiv, Google Scholar, DOI.org) and interacts with the official Zotero API/MCP for legitimate research management purposes. These references are considered safe and consistent with the skill's primary purpose.

Research Ideation