The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches financial data in JSON format from https://hhxg.top/static/data/. This communication is the primary function of the skill and targets the author's own domain.
[COMMAND_EXECUTION]: The skill includes a utility gen_screenshots.py that uses subprocess.run to execute other scripts within the package to generate documentation assets. The skill instructions also guide the agent to use the Bash tool to run the data fetching scripts.
[PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection by processing external data and passing it to the agent context.
Ingestion points: Remote JSON data is retrieved via urllib.request in scripts/_common.py.
Boundary markers: No explicit delimiters or ignore-instructions are used when presenting external data to the agent.
Capability inventory: The agent is granted Bash access to run internal scripts.
Sanitization: The skill formats API responses into markdown tables and summaries without specific escaping for potential instructions embedded in the market data.

hhxg-market