The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill retrieves documentation and API specifications from the official vendor domain (www.mux.com) and official GitHub repositories (github.com/muxinc/). These downloads are used to provide the agent with current technical information.
[COMMAND_EXECUTION]: The instructions direct the agent to use the mux CLI (specifically mux docs find) to search for relevant documentation topics locally. This is a scoped use of a vendor-provided tool.
[INDIRECT_PROMPT_INJECTION]: The skill processes external markdown content and JSON specifications from Mux's documentation site. While reading external data is a known surface for indirect prompt injection, the sources are restricted to official vendor-controlled domains and the activity is central to the skill's primary purpose.
Ingestion points: Documentation pages (.md), collection indexes (.txt), and API specifications (.json) from www.mux.com and github.com/muxinc/.
Boundary markers: Not explicitly defined in the instructions.
Capability inventory: File reading and network fetching via standard agent tools; execution of the mux CLI tool.
Sanitization: Not explicitly mentioned; the skill relies on the structured nature of the documentation.

mux-video