The Agent Skills Directory

[PROMPT_INJECTION]: The skill establishes internal operational protocols (e.g., BUILD vs EXPERIMENT modes) for agent behavior. There are no attempts to bypass safety filters, extract system prompts, or induce unrestricted behavior patterns.- [DATA_EXFILTRATION]: No sensitive data access patterns or network exfiltration methods were identified. The framework actually advocates for security hygiene by recommending the redaction of secrets and PII before sharing context.- [COMMAND_EXECUTION]: No shell commands, terminal invocations, or dynamic command injection patterns (such as !backtick syntax) are present in the skill content.- [NO_CODE]: The skill consists exclusively of markdown instructions and metadata without any accompanying scripts, binaries, or automated tool definitions.

pricing-strategy