The Agent Skills Directory

[SAFE]: The skill's instructions and reference guides are entirely focused on academic writing best practices and do not contain prompt injection or bypass attempts.
[SAFE]: The utility script scripts/check_paper.py is a standalone tool for stylistic analysis. It uses standard Python libraries (re, sys, pathlib, collections), requires no external dependencies, and lacks network or sensitive file access capabilities.
[INDIRECT_PROMPT_INJECTION]: The skill involves processing user-authored academic text. While this constitutes a data ingestion surface, the skill is focused on stylistic and structural advice and does not possess capabilities that could be maliciously triggered by content within a research paper. 1. Ingestion points: User-provided files are read by the agent and the scripts/check_paper.py script. 2. Boundary markers: None identified in the skill instructions. 3. Capability inventory: The skill uses standard file reading capabilities; no dangerous subprocesses, dynamic code execution, or network operations are present. 4. Sanitization: No input sanitization is performed on the user text, as the focus is on stylistic analysis.

paper-writing