The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the TikHub API, which reflects user-generated content (titles, descriptions, user nicknames) from Xiaohongshu. This content is processed and saved to local files or printed to the console without sanitization, creating a surface where malicious instructions embedded in social media posts could influence the agent's behavior.
Ingestion points: API responses from api.tikhub.io are read in the Python snippets and the postprocess_xiaohongshu_raw.py script.
Boundary markers: The skill does not implement delimiters or instructions to ignore potential commands within the fetched data.
Capability inventory: The skill scripts utilize httpx for network communication and standard file I/O for writing JSON and CSV files.
Sanitization: There is no evidence of filtering or escaping logic for the fields extracted from the API (e.g., note_info.title, note_info.desc).
[EXTERNAL_DOWNLOADS]: The skill is designed to communicate with an external third-party API at api.tikhub.io. It sends a user-provided API key (TIKHUB_API_KEY) to this domain. While this is the primary purpose of the skill, users should ensure they trust the endpoint before providing credentials.
[COMMAND_EXECUTION]: The instructions include several shell commands for the user to execute, including pip install httpx, curl requests, and running a provided Python post-processing script that performs file system operations.

tikhub-xiaohongshu-search