The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to scrape and process content from various external sources including LinkedIn, GitHub, X/Twitter, YouTube, and generic web pages. This creates a significant surface for indirect prompt injection attacks where malicious instructions hidden in the scraped content could be interpreted and executed by the agent.
Ingestion points: Scrape endpoints defined in references/detailed-guide.md for websites, social media profiles, and search results.
Boundary markers: The instructions lack explicit delimitation or "ignore embedded instructions" directives for the data retrieved from external sources.
Capability inventory: The skill possesses the ability to perform network requests (HTTP POST/GET) and manage email workflows (drafting and sending), providing a path for potential automated exploitation if the agent is compromised by injected content.
Sanitization: No sanitization or validation of the ingested external content is specified before it is processed by the agent.
[DATA_EXFILTRATION]: The skill performs network operations to external endpoints defined by the DEEPAPI_API_BASE_URL environment variable and references the deepapi.co domain for administrative tasks.
Network operations: The skill initiates HTTP requests to a non-whitelisted external domain to fulfill its scraping and research functions. While this is the intended purpose of the skill, it represents a network communication channel to a third-party service.

deepapi