The Agent Skills Directory

COMMAND_EXECUTION (LOW): The configuration utility in scripts/request/web/encrypt/config.py uses the eval() function to parse variables from the web_encrypt_config.ini file. This is primarily used to interpret structured data like integer lists. While the configuration source is local to the skill, the use of eval() is a security best-practice violation.
PROMPT_INJECTION (LOW): The skill is susceptible to indirect prompt injection because it retrieves and processes untrusted content (notes and comments) from external social media users.
Ingestion points: Data enters through functions in scripts/request/web/apis/note.py and scripts/request/web/apis/comments.py.
Boundary markers: The skill does not use delimiters or boundary markers when returning retrieved content to the agent.
Capability inventory: The skill possesses authenticated interaction capabilities, such as following users and liking notes, which could be exploited by malicious instructions in scraped data.
Sanitization: Content fetched from the platform is not sanitized or escaped.
EXTERNAL_DOWNLOADS (SAFE): The skill relies on standard, legitimate Python packages for networking and cryptography, including aiohttp, loguru, pycryptodome, and getuseragent.

xiaohongshu