The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill's scripts (rg_rank.py, make_code.py, and sdk_info.py) fetch API swagger definitions, search results, and SDK metadata from official Volcengine API endpoints (api.volcengine.com). This is the intended functionality of the skill to provide up-to-date documentation and code samples.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external Volcengine API endpoints to generate code samples and search results, which represents a potential attack surface for indirect prompt injection if the remote data source were compromised.
Ingestion points: API search results and swagger documentation fetched from https://api.volcengine.com via urllib.request.
Boundary markers: Not explicitly defined for the external data in the system prompts.
Capability inventory: The skill generates and displays code samples to the user; it does not automatically execute the generated code or perform sensitive file/network operations based on the external input.
Sanitization: The scripts include functions like clean_highlight to remove HTML tags from search results and scrub_unusable_json_examples to sanitize example values in generated code.

volcengine-sdk-generator