The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructs the agent to gather and analyze data from untrusted external sources, such as Reddit, Quora, and competitor websites. This creates a surface for indirect prompt injection, where malicious instructions embedded in that external content could influence the agent's behavior during analysis. * Ingestion points: Forum research and competitor analysis sources identified in SKILL.md. * Boundary markers: The instructions lack explicit delimiters or warnings to treat external data as untrusted content. * Capability inventory: The skill focuses on strategic planning and does not include scripts for code execution, system modification, or network exfiltration. * Sanitization: No logic is provided to sanitize or escape data retrieved from external research sources.

content-strategy