The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions to download and run code from a third-party repository (github.com/magnum6actual/flipoff). This source is not affiliated with a recognized organization.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data for display, creating an attack surface for indirect prompt injection.
Ingestion points: Data from the MESSAGES array in js/constants.js and external data sources such as the /api/departures endpoint shown in implementation examples in SKILL.md.
Boundary markers: No delimiters or boundary instructions are present in the provided snippets to separate untrusted data from the agent's processing context.
Capability inventory: Capabilities described in SKILL.md and related files like Board.js and SoundEngine.js are restricted to standard DOM manipulation and audio processing; no sensitive capabilities like system shell execution or file system writes are exposed.
Sanitization: External strings are converted to uppercase for display but are not otherwise validated, escaped, or filtered to prevent injection.

flipoff-split-flap-display