The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions and guides the agent to execute npx commands for its own installation and for the optional installation of supporting tools such as web-search and competitor-teardown.
[EXTERNAL_DOWNLOADS]: The skill fetches its code and additional research capabilities from the inference-sh/skills repository. These downloads are associated with the skill's official distribution channel.
[PROMPT_INJECTION]: The skill processes untrusted information retrieved from the open web (via the web-search tool) to populate report sections and message templates. This constitutes an indirect prompt injection surface.
Ingestion points: Data retrieved from external websites, LinkedIn, and Instagram during research rounds 1-6 enters the agent context to be processed into the final document.
Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are used when processing the external search results.
Capability inventory: The skill utilizes the Write tool to create persistent .md and .html files in the current working directory.
Sanitization: No explicit sanitization or escaping logic is defined for the interpolation of external data into the generated output files.
[SAFE]: The skill's primary functionality—researching companies and generating reports—is performed using legitimate tools. It does not attempt to access sensitive system files, hardcode credentials, or establish unauthorized network connections beyond the research queries.

prospect-360-b2b