mongodb-schema-design
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external database schema data and query results via the MongoDB MCP integration, creating a potential surface for indirect prompt injection.\n
- Ingestion points: External data enters the agent context through tools such as
mcp__mongodb__collection-schema,mcp__mongodb__aggregate, andmcp__mongodb__db-statsreferenced inSKILL.md.\n - Boundary markers: The skill includes a mandatory "Action Policy" in
SKILL.mdthat requires the agent to summarize operations and seek explicit user confirmation before executing any write or destructive commands.\n - Capability inventory: The skill environment includes high-privilege capabilities for data modification and deletion via tools like
update-many,insert-many,delete-many,drop-collection, anddrop-database.\n - Sanitization: While procedural guardrails (confirmation steps) are defined, the skill does not explicitly define technical sanitization or filtering for ingested database content.\n- [EXTERNAL_DOWNLOADS]: The skill references the official MongoDB MCP Server for database analysis tasks.\n
- Evidence:
SKILL.mdprovides a reference link tohttps://github.com/mongodb-js/mongodb-mcp-server.\n - Context: The link points to a repository maintained by a well-known service (MongoDB) and is intended for legitimate database integration purposes.
Audit Metadata