The Agent Skills Directory

[Prompt Injection] (SAFE): The instructions are focused on presentation workflows. There are no patterns suggesting attempts to bypass safety filters or override core agent instructions.
[Data Exposure & Exfiltration] (SAFE): The skill does not access sensitive system paths (like ~/.ssh or ~/.aws) and contains no network-capable commands (curl, wget, etc.).
[Obfuscation] (SAFE): No encoded strings (Base64), zero-width characters, or homoglyphs were detected in the markdown files.
[Unverifiable Dependencies & Remote Code Execution] (SAFE): The skill does not define any external dependencies (npm, pip) and contains no scripts or remote code execution patterns.
[Indirect Prompt Injection] (LOW):
Ingestion points: Untrusted data enters via user-provided content (docs, bullets, metrics) as defined in SKILL.md and references/INTAKE.md.
Boundary markers: The skill does not specify markers to delimit user data.
Capability inventory: The skill has no capabilities beyond generating markdown text (no subprocesses, no file-writes, no network ops).
Sanitization: No sanitization is performed on user inputs.
Assessment: The vulnerability surface is present but lacks any exploitable capabilities, resulting in a low-risk profile.
[No Code] (INFO): The skill contains no executable scripts (Python, Node.js, etc.); it is comprised solely of Markdown instructions and configuration.

giving-presentations