The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions to download the Apollo MCP Server binary and installer from the official vendor domain (mcp.apollo.dev). These downloads are associated with the primary purpose of the skill and originate from the vendor's infrastructure.\n- [REMOTE_CODE_EXECUTION]: The quick start guide contains installation patterns that pipe remote scripts directly to a shell (curl | sh for Unix-based systems and iwr | iex for Windows). These are documented as the official installation methods for the tool.\n- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to install, configure, and start the MCP server, as well as to use the Apollo Rover CLI and connect via mcp-remote.\n- [INDIRECT_PROMPT_INJECTION]: The skill involves processing external GraphQL schemas and operation files, which serves as a potential attack surface for indirect prompt injection.\n
Ingestion points: Local GraphQL schema files (schema.graphql) and operation files (stored in ./operations/) are read and processed by the agent's introspection tools.\n
Boundary markers: No specific delimiters or instructions to ignore embedded content were identified in the prompt logic.\n
Capability inventory: The agent has access to file system operations (Read, Write, Edit, Glob) and shell execution (Bash).\n
Sanitization: The skill does not explicitly describe sanitization or filtering of the schema or operation content before it is processed by the agent.\n- [DYNAMIC_EXECUTION]: The skill documents the use of Rhai scripting, an embedded language that allows the server to execute custom logic at runtime. The documentation notes that these scripts can read environment variables and make outbound HTTP requests.

apollo-mcp-server