The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to fetch and process data from external sources using searching tools, which introduces a surface for processing untrusted content.
- Ingestion points: External market, company, and industry evidence are retrieved via tavily_search and exa_search tool calls (references/sandbase-api-map.md).
- Boundary markers: The skill lacks specific instructions for the agent to treat retrieved web content as untrusted or to use distinct delimiters to isolate external text from internal reasoning.
- Capability inventory: The agent utilizes sandbase_call_tool to perform network-based searches and sandbase_describe_tool to interact with tool schemas (references/sandbase-api-map.md).
- Sanitization: There are no defined procedures for filtering, escaping, or validating the content returned from external search providers before the agent incorporates it into its context.

market-sizing-analysis