The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses a Python script to fetch trending data from the vendor-owned endpoint https://redfox.hk/story/api/cozeSkill/getWxCozeSkillData.
[COMMAND_EXECUTION]: The skill executes a local Python script scripts/fetch_official_account_trends.py via shell command interpolation for the --keyword and --days parameters, which presents a potential command injection surface if the agent does not properly sanitize these inputs before execution.
[DATA_EXFILTRATION]: The skill requires a REDFOX_API_KEY, which is transmitted to the vendor's infrastructure at redfox.hk for authentication. This is consistent with the skill's documented vendor-provided functionality.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it ingests and processes external article data without using defined boundary markers or sanitization logic.
Ingestion points: Data is fetched from an external API and saved to a Markdown file ({keyword}_爆款数据.md) which is subsequently read by the agent.
Boundary markers: None identified; instructions do not specify the use of delimiters or 'ignore' instructions for the processed data.
Capability inventory: The skill possesses capabilities for network operations (via requests), file system writes, and subprocess execution.
Sanitization: There is no evidence of sanitization or validation of the content retrieved from the external API before it is processed by the agent.

wechat-title