The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructions direct the user to configure an external MCP server at https://rube.app/mcp. This endpoint provides the tool schemas and execution environment for the Pipedrive automation logic.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from Pipedrive CRM, which creates a potential surface for indirect prompt injection if records contain malicious instructions.
Ingestion points: Data is pulled into the agent's context through tools such as PIPEDRIVE_SEARCH_PERSONS, PIPEDRIVE_GET_DETAILS_OF_A_PERSON, PIPEDRIVE_GET_ALL_NOTES, and PIPEDRIVE_GET_ALL_ACTIVITIES_ASSIGNED_TO_A_PARTICULAR_USER as defined in SKILL.md.
Boundary markers: None identified. There are no instructions to the agent to treat data from these tools as untrusted or to ignore instructions embedded within the record fields.
Capability inventory: The skill possesses capabilities to write and modify CRM data, including PIPEDRIVE_ADD_A_DEAL, PIPEDRIVE_ADD_A_NOTE, PIPEDRIVE_UPDATE_A_PERSON, and PIPEDRIVE_ADD_AN_ACTIVITY as documented in SKILL.md.
Sanitization: While the documentation mentions that HTML content in notes is sanitized server-side for formatting, there is no mention of sanitizing or filtering natural language instructions to prevent the agent from executing them.

pipedrive-automation