The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from external business websites and SEO research tools to generate automated teardowns and cold outreach emails. This creates a surface where malicious instructions embedded in a target website could influence the agent's behavior during email drafting or report generation.
Ingestion points: The research_lead and get_lead_teardown tools ingest external web content, site crawl data, and competitor profile information (found in SKILL.md under 'HarborRank MCP tools').
Boundary markers: The instructions do not specify the use of delimiters or explicit 'ignore embedded instructions' warnings when processing external site data.
Capability inventory: The skill possesses the ability to write files to the local file system (outreach/leads/), make network requests via MCP tools, and generate outreach communications.
Sanitization: There is no mention of sanitizing, escaping, or filtering the external content before it is interpolated into prompts or generated documents.
[EXTERNAL_DOWNLOADS]: The skill relies on an external Model Context Protocol (MCP) server located at https://app.harborrank.com/mcp to provide its core SEO research and lead management capabilities. This domain corresponds to the vendor's own infrastructure.
[COMMAND_EXECUTION]: The documentation provides users with specific shell commands to manually configure the MCP environment, such as claude mcp add --transport http --scope user harborrank https://app.harborrank.com/mcp. These are setup instructions provided to the user rather than automated execution by the skill itself.

lead-teardown