The Agent Skills Directory

[SAFE]: The skill is designed to improve academic research integrity by verifying the existence and accuracy of citations. All instructions and scripts align with this primary purpose.
[EXTERNAL_DOWNLOADS]: The skill references standard, well-known scholarly APIs such as CrossRef, arXiv, and Semantic Scholar. It also identifies legitimate Python dependencies (requests, bibtexparser, arxiv, semanticscholar) required for its reference scripts. These are documented neutrally as they originate from trusted or well-known scholarly services.
[DATA_EXFILTRATION]: Network operations are restricted to GET requests against established academic metadata repositories. The skill does not access sensitive local files (e.g., SSH keys, credentials) or attempt to transmit private user data to unknown third-party servers.
[INDIRECT_PROMPT_INJECTION]: While the skill ingests potentially untrusted BibTeX and LaTeX content, the threat is minimized. The processing logic is limited to metadata extraction and matching, uses regex-based normalization, and does not involve executing the input data or interpolating it into high-privilege command contexts.
[PROMPT_INJECTION]: The instructional content contains natural markers such as 'IMPORTANT' and 'Core Principle', but these are used for providing legitimate guidance on citation verification and do not attempt to override the underlying agent's safety protocols or system instructions.

citation-verification