The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources (Twitter, TikTok, YouTube, Google News, Reddit, Weibo) via the sandbase_call_tool interface. This content could theoretically contain instructions designed to influence the agent's behavior. However, the skill includes guidelines for read-only research and source attribution to mitigate these effects.
Ingestion points: External data is retrieved from social media and news platforms via sandbase_call_tool and processed within the agent context (SKILL.md, references/sandbase-api-map.md).
Boundary markers: No explicit delimiters or boundary markers are defined for isolating the external data content from the agent's instructions.
Capability inventory: The skill uses sandbase_describe_tool and sandbase_call_tool to interact with the external platform APIs.
Sanitization: No specific sanitization, filtering, or validation of the external content is mentioned before it is synthesized into research findings.

trend-spotter