The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted content from social media (KOL tweets) and news feeds. Ingestion points: SKILL.md contains multiple curl calls to endpoints like /tweets/feed and /info/feed which retrieve external data. Boundary markers: There are no explicit boundary markers or instructions to ignore embedded commands within the fetched data in the provided prompt templates. Capability inventory: The skill has access to curl and jq in SKILL.md to interact with the vendor's API. Sanitization: The instructions do not specify any sanitization or filtering of the retrieved text before it is used in AI synthesis.
[COMMAND_EXECUTION]: The skill executes the curl and jq binaries to fetch and process market data from the vendor's API. Evidence of this is found in the multiple curl commands defined in SKILL.md.
[EXTERNAL_DOWNLOADS]: The skill downloads data from the vendor's domain https://api.ctmon.xyz/api to provide its market monitoring services, as documented in the configuration section of SKILL.md.

ct-monitor