The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill retrieves and reads external demo source code via the mcp__pmndrs__get_example tool and examples://index resource. This creates an attack surface where untrusted content could contain instructions to influence the agent's behavior.
Ingestion points: ReadMcpResourceTool (index) and mcp__pmndrs__get_example (source code).
Boundary markers: No delimiters or specific instructions to ignore embedded commands are specified when processing the external content.
Capability inventory: The agent analyzes and explains the retrieved source code to the user.
Sanitization: No sanitization is performed on the source code retrieved from the external repository.
[COMMAND_EXECUTION]: The skill provides the agent with a template command to provide to users for downloading project assets.
Evidence: npx degit pmndrs/examples/examples/{name} in SKILL.md.
Note: The command is intended for manual execution by the user and refers to the author's pmndrs/examples repository.

examples