The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides commands to fetch installation scripts from https://rover.apollo.dev/nix/latest and https://rover.apollo.dev/win/latest. These are official distribution endpoints for the Apollo Rover toolset.
[REMOTE_CODE_EXECUTION]: The provided installation workflow uses shell piping to execute remote content (curl | sh and iwr | iex). This is a high-capability pattern, though documented here for legitimate software setup.
[INDIRECT_PROMPT_INJECTION]: The skill functions by fetching and analyzing GraphQL schemas from external sources, creating a surface for indirect prompt injection attacks.
Ingestion points: External data enters the agent context via rover graph fetch, rover graph introspect, and rover subgraph introspect (SKILL.md, references/graphs.md, references/subgraphs.md).
Boundary markers: No specific delimiters or instructions are provided to the agent to treat external schema SDL as untrusted content.
Capability inventory: The skill is configured to use the Bash tool for running CLI commands and has permissions for reading and writing files (SKILL.md).
Sanitization: The skill lacks validation or sanitization mechanisms to filter malicious instructions potentially embedded in GraphQL schema comments or metadata.

rover