The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The script scripts/fetch_xhs_trends.py connects to an external API endpoint at https://onetotenvip.com/skill/cozeSkill/getXhsCozeSkillData to fetch trending data.
[COMMAND_EXECUTION]: The Python script scripts/fetch_xhs_trends.py utilizes a custom network handler that deliberately bypasses standard security protocols.
Insecure SSL Configuration: The script explicitly sets ssl.verify_mode = ssl.CERT_NONE and context.check_hostname = False. This disables SSL certificate validation, rendering the connection vulnerable to Man-in-the-Middle (MitM) attacks where data could be intercepted or altered.
SNI Suppression: The script uses server_hostname=None to suppress Server Name Indication (SNI). This is an unusual practice often associated with evading network traffic filters or analysis tools.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection because it ingests and processes data from an external API.
Ingestion points: External content is retrieved via scripts/fetch_xhs_trends.py and passed to the agent to generate summaries and recommendation reasons.
Boundary markers: The instructions do not define clear delimiters or "ignore previous instructions" guards for the data returned from the API.
Capability inventory: The skill has the capability to execute scripts and write files to the local file system (.html output).
Sanitization: There is no evidence of sanitization or validation of the fetched content before it is processed by the agent, meaning malicious instructions embedded in the API response could potentially influence agent behavior.

xhs-explosive-content-detector