The Agent Skills Directory

[SAFE]: The skill serves its primary purpose of documenting Intercom automation through the Rube MCP server. No malicious patterns such as credential exfiltration, remote code execution, or obfuscation were detected.
[PROMPT_INJECTION]: While the skill processes external content (Intercom conversation messages) which is a potential surface for indirect prompt injection, this is inherent to its functionality as a customer support automation tool.
Ingestion points: SKILL.md (via tools like INTERCOM_GET_CONVERSATION and INTERCOM_LIST_CONVERSATIONS which retrieve untrusted message bodies).
Boundary markers: Absent.
Capability inventory: SKILL.md (includes tools for acting on the platform: INTERCOM_REPLY_TO_CONVERSATION, INTERCOM_ASSIGN_CONVERSATION, INTERCOM_CLOSE_CONVERSATION).
Sanitization: Present (the documentation explicitly advises sanitizing HTML input to prevent rendering issues).

intercom-automation