joinlayer-pipelines
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to configure and connect to the JoinLayer MCP server using the vendor's official endpoint.
- Details: The agent is directed to use the deployment URL at https://mcp.joinlayer.app/mcp for MCP tool discovery and authentication.
- [INDIRECT_PROMPT_INJECTION]: The skill defines workflows for processing data through various connectors, creating a surface where instructions could theoretically be embedded in source data.
- Ingestion points: Data is ingested from external source connections, such as Kafka topics and database tables, into the pipeline environment.
- Boundary markers: The skill explicitly instructs the agent to separate durable layers in reasoning, verify identity and workspace roles first, and obtain user confirmation for material impacts.
- Capability inventory: MCP tools allow the agent to read schemas, update pipeline configurations, and start or stop data movement runs.
- Sanitization: Mandatory validation and preview steps are required to inspect transformed target rows and resolve blockers before execution.
Audit Metadata