The Agent Skills Directory

[Prompt Injection] (SAFE): The identity and usage instructions in SKILL.md are focused on grounding the model in specific reference files to ensure technical consistency and do not contain instructions to override safety constraints or reveal system prompts.
[Data Exposure & Exfiltration] (SAFE): No hardcoded credentials or sensitive file paths were detected. The security keywords (e.g., API_KEY, SECRET) found in references/validations.md are part of diagnostic regex patterns intended to flag potential security issues in user code.
[Obfuscation] (SAFE): No Base64, zero-width characters, or other encoding techniques were found in the skill content.
[Unverifiable Dependencies & Remote Code Execution] (SAFE): The skill does not perform any package installations or remote code downloads. Mentioned packages like @sveltejs/kit and standard adapters are used as context for architectural advice.
[Privilege Escalation] (SAFE): No commands involving sudo or permission modification were identified.
[Indirect Prompt Injection] (SAFE): The skill has a low attack surface as it contains no executable code or tools. 1. Ingestion points: User-provided SvelteKit code. 2. Boundary markers: Explicit grounding in reference files. 3. Capability inventory: None (no scripts folder). 4. Sanitization: N/A.
[Dynamic Execution] (SAFE): No dynamic execution patterns (eval, exec) or runtime compilation were detected.

sveltekit