The Agent Skills Directory

[SAFE]: The skill is primarily instructional, providing a detailed persona and structured workflow for summarizing academic papers using life metaphors and anime references. No executable scripts are included.
[PROMPT_INJECTION]: The skill uses detailed role-play instructions to establish a specific persona ('Grandma'). This is a legitimate use of instructions for customizing agent behavior and does not contain attempts to bypass safety guardrails or ignore system prompts.
[DATA_EXPOSURE]: The skill instructions involve reading local PDF files and fetching content from arXiv URLs. These are standard operations for a paper summarization tool and do not involve unauthorized exfiltration of sensitive data or credentials.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it is designed to ingest and process untrusted data from external academic papers. However, the instructions are limited to summarization and storytelling, and the skill does not request access to high-privilege tools that could be abused via malicious content in processed papers.
Ingestion points: Local PDF files, arXiv URLs, and pasted text specified in SKILL.md.
Boundary markers: Absent.
Capability inventory: None. The skill provides instructions for interpretation and storytelling but does not include scripts or tools for file modification or arbitrary command execution.
Sanitization: Absent.

paper-reading