The Agent Skills Directory

[SAFE]: The skill acts as an informational resource for Superhuman users, providing pricing details, feature comparisons, and troubleshooting steps.
[EXTERNAL_DOWNLOADS]: Documents the setup of the official Superhuman MCP server using the @anthropic-ai/claude-code-mcp-superhuman package. This is a well-known service provided by a reputable organization.
[COMMAND_EXECUTION]: Provides instructions for recommending and installing related skills from the same vendor (sales-skills) using standard platform commands. This is consistent with the skill's role as a platform navigator.
[PROMPT_INJECTION]: Implements a 'learnings' mechanism that reads from and writes to references/learnings.md. While this creates an indirect prompt injection surface for persisting information across sessions, it is an intended feature of the skill's persistent learning capability.
Ingestion points: Data is loaded from references/learnings.md at the start of the workflow.
Boundary markers: None identified.
Capability inventory: The skill provides platform advice and navigates to other vendor skills.
Sanitization: None identified.

sales-superhuman