The Agent Skills Directory

[SAFE]: The skill leverages authorized vendor tools (Celigo CLI and MCP) and follows established platform schemas (lookups.yml, mappings.yml, transform.yml) for data transformation tasks. It correctly directs users to manage secrets using account settings rather than hardcoding credentials.
[INDIRECT_PROMPT_INJECTION]: The skill describes processes that involve reading live source records and error retry data to facilitate mapping creation, which establishes an indirect prompt injection attack surface.
Ingestion points: Untrusted record data enters the agent context via celigo exports invoke (references/cli.md) and get_flow_error_retry_data (references/mcp.md).
Boundary markers: The instructions lack specific guidance on using delimiters or protective instructions to isolate ingested data from the agent's core logic.
Capability inventory: The skill provides the agent with write capabilities to the Celigo platform using upsert_import, upsert_export, and upsert_flow (references/mcp.md).
Sanitization: No explicit content sanitization or validation routines are described for the data ingested from external integration steps.

writing-mappings