The Agent Skills Directory

[SAFE]: The skill focuses on generating Markdown content for HackMD presentations and does not exhibit malicious behavior. It includes clear warnings to avoid executable code and custom plugins.
[EXTERNAL_DOWNLOADS]: The skill references documentation and public repositories from well-known services, specifically HackMD and the reveal.js GitHub repository, for legitimate documentation and syntax reference.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided source material to generate slide decks. While this involves handling untrusted data, the instructions include explicit constraints to prevent the inclusion of executable code or unauthorized plugins in the output.
Ingestion points: Processes user-requested topics and source material in SKILL.md and references/syntax.md.
Boundary markers: Recommends using long code fences to enclose Markdown output in SKILL.md.
Capability inventory: Uses external tools/CLI to interact with HackMD notes, but explicitly restricts these to content authoring rather than account operations.
Sanitization: Strictly prohibits the use of plugins, dependencies, and JavaScript callbacks in SKILL.md and references/syntax.md.

hackmd-slide