The Agent Skills Directory

[DATA_EXFILTRATION]: The skill communicates with i.weread.qq.com, which is the official API gateway for the WeRead service owned by Tencent. It uses a WEREAD_TOKEN (API key) for authentication, which is standard for such integrations. The token is obtained via the user's account and used to fetch personal reading data.
[COMMAND_EXECUTION]: The documentation provides examples using curl and mentions a troubleshooting utility zero doctor. These are used for testing connectivity and verifying the environment, which is consistent with the skill's developer-focused documentation.
[PROMPT_INJECTION]: The skill processes untrusted external data in the form of book reviews (/review/list) and search results (/store/search). This represents a potential indirect prompt injection surface.
Ingestion points: External book titles, author names, and public reader reviews are ingested into the agent's context through several sub-API calls.
Boundary markers: None are specified in the current instructions to separate external data from system instructions.
Capability inventory: The skill uses network requests to the WeRead API but does not show evidence of dangerous local file system or arbitrary code execution capabilities.
Sanitization: No explicit sanitization of the retrieved text is mentioned. However, this risk is inherent to the primary function of the skill (reading books and reviews) and is considered acceptable within this context.

weread