The Agent Skills Directory

[SAFE]: The skill provides guidelines and examples for formatting citations in various styles (APA, MLA, etc.) using natural language instructions. It does not include any scripts, executable files, or network-enabled tools.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted user data in the form of citation strings and research metadata.
Ingestion points: User citation data provided in SKILL.md examples and instructions.
Boundary markers: None defined; the skill lacks specific delimiters to separate user data from instructions.
Capability inventory: No capabilities detected. The skill does not request tool access in its frontmatter and contains no code for file system or network operations.
Sanitization: Not applicable as the skill possesses no exploitable capabilities.
[METADATA_POISONING]: The skill attributes itself to 'Claude Code' in the YAML frontmatter while linking to the 'inbharatai' repository in the footer. This appears to be a branding choice rather than a deceptive practice intended to bypass security controls.

citation-manager