The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates the retrieval and processing of data from external SaaS applications (such as Slack, Jira, Notion, and HubSpot) through the execute_tool function. This creates an attack surface where malicious instructions hidden within third-party data could influence the agent's actions.\n- Ingestion points: Data enters the agent's context through execute_tool calls to live third-party APIs, as described in SKILL.md and references/tool-discovery.md.\n- Boundary markers: The instructions do not define delimiters or specific boundary markers to separate external data from internal instructions, although user confirmation for write operations is advised.\n- Capability inventory: The skill possesses significant capabilities, including the ability to perform read and write operations on SaaS accounts and administrative management of the Elaichi organization (members, roles, and teams) through elaichi__* tools.\n- Sanitization: The documentation does not specify any sanitization, filtering, or validation processes for the content returned by the external tools.

elaichi-mcp