The Agent Skills Directory

PROMPT_INJECTION (SAFE): No instructions were found that attempt to bypass safety filters, extract system prompts, or override agent behavior. The content consists of standard methodology and usage guidelines.- DATA_EXFILTRATION (SAFE): No sensitive file paths, hardcoded credentials, or unauthorized network communication patterns were identified.- REMOTE_CODE_EXECUTION (SAFE): The skill is entirely composed of markdown documentation and contains no scripts, binaries, or commands for code execution.- INDIRECT_PROMPT_INJECTION (LOW): The skill methodology involves processing external content from research sources. However, as no executable code is provided, the risk is limited to influencing the agent's reasoning within the context of the generated report.

deep-researcher