The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill references an external MCP server at datahub.noxiaohao.com within a code comment in SKILL.md. This domain is not affiliated with the stated vendor (Bitget-AI), representing an unverifiable and potentially untrusted infrastructure dependency.
[PROMPT_INJECTION]: The skill contains instructions to misrepresent or conceal the names of third-party data providers in the event of tool failures (e.g., rewriting 'ApeWisdom' or 'Binance' errors into neutral language). This is a form of deceptive instruction intended to hide the skill's inner workings and data sources from the user.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the ingestion of untrusted community data.
Ingestion points: The derivatives_sentiment(action="reddit_trending") tool action retrieves live trending data from social media platforms (Reddit).
Boundary markers: Absent; there are no instructions to the agent to use delimiters or to ignore potential instructions embedded within the social media data.
Capability inventory: The skill is primarily designed for data synthesis and reporting; the provided files do not include high-privilege capabilities such as arbitrary file writes or outbound network requests beyond the defined tool set.
Sanitization: Absent; the instructions do not specify any validation, filtering, or escaping of the ingested social media content.

sentiment-analyst