The Agent Skills Directory

[PROMPT_INJECTION]: The skill contains no instructions aimed at bypassing AI safety guardrails or extracting system prompts. It provides a structured framework for generating prompts for other agents while emphasizing grounding and accuracy.
[DATA_EXFILTRATION]: There are no network operations, API calls, or file-read operations that could lead to data exfiltration. The skill is purely text-transformative.
[CREDENTIALS_UNSAFE]: No hardcoded credentials, tokens, or private keys were found. The use of double curly braces {{like_this}} is correctly identified as a placeholder for user-supplied data in generated prompts.
[COMMAND_EXECUTION]: The skill does not contain any shell commands, subprocess calls, or scripts that interact with the host environment.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest user-supplied tasks or rough prompts. While this is an ingestion surface, the skill implements strict 'Grounding' rules to ensure the generated output only uses verified facts and explicitly states assumptions, which mitigates the risk of the model following malicious instructions embedded in the input data.
[OBFUSCATION]: The content is entirely in plain-text markdown with no hidden characters, encoded strings, or homoglyph attacks.

prompt-generator