ToolJet
An open-source low-code application development platform that supports the MCP protocol, allowing coding agents to directly operate workspaces, create data tables, and build interface components.
Core Facts
Timeline (last 90 days)
ToolJet applies RBAC at two levels: platform permissions determine who can reach a given workspace resource; application permissions determine which pages/queries/components each group can open and run.
The principle of separating build time from runtime: at build time the coding agent, via MCP, completes table schema creation, query wiring, page layout, and permission configuration in one go, producing a static self-contained application; at runtime the application executes independently relying only on the already-configured data sources, requiring no agent online, consuming no model tokens, and unaffected by the agent session lifecycle.
Separating build time from runtime brings three benefits: the application can run stably over the long term, costs are predictable (each analysis pays only for the Jev calls), and security is better (runtime does not require granting the agent workspace write permissions).
Once requirements change, an application generated using the build-time/runtime separation principle needs to be regenerated or manually adjusted, and will not automatically follow prompt changes.
In an application built with ToolJet, a viewer calling a write query receives a 401 rejection, indicating that the server is indeed enforcing RBAC rules.
The official ToolJet blog showcases a 'Support Ticket Sorter' app generated by a coding agent, using ToolJet MCP to build the app skeleton and TypeSafe's Jev to make the sorting judgments, with 100 synthetic tickets ultimately turned into a reviewable processing queue.
ToolJet MCP is an implementation of this protocol on the ToolJet low-code platform, enabling coding agents to directly read the workspace structure, create data tables and queries, and generate interface components.
The app uses three ToolJet groups, Support Sorter - Agents, Leads, and Viewers, to divide responsibilities: only Agents and admins can execute sorting, Jev, save, and note operations, Reset is restricted to admins, and Leads and Viewers are read-only; the rules are enforced at the query level, and blocked users receive a server-side rejection.
ToolJet provides governance capabilities at the platform layer, supporting SAML, OIDC, or LDAP single sign-on with SCIM automated account provisioning, role-based access control refined down to apps, data sources, and each query, audit logs recording every login, edit, and approval decision, Docker or Kubernetes self-hosted air-gapped deployment, multi-user collaborative editing with version management and Git sync, and ToolJet AI can run within its own deployment environment.
After switching to Plan B, scene loading time was reduced from 3.3 seconds to approximately 0.5 seconds, approximately 6.8 times faster.
All Facts (10)
After switching to Plan B, scene loading time was reduced from 3.3 seconds to approximately 0.5 seconds, approximately 6.8 times faster.
75%UnverifiedToolJet applies RBAC at two levels: platform permissions determine who can reach a given workspace resource; application permissions determine which pages/queries/components each group can open and run.
50%UnverifiedThe principle of separating build time from runtime: at build time the coding agent, via MCP, completes table schema creation, query wiring, page layout, and permission configuration in one go, producing a static self-contained application; at runtime the application executes independently relying only on the already-configured data sources, requiring no agent online, consuming no model tokens, and unaffected by the agent session lifecycle.
50%UnverifiedSeparating build time from runtime brings three benefits: the application can run stably over the long term, costs are predictable (each analysis pays only for the Jev calls), and security is better (runtime does not require granting the agent workspace write permissions).
50%UnverifiedOnce requirements change, an application generated using the build-time/runtime separation principle needs to be regenerated or manually adjusted, and will not automatically follow prompt changes.
50%UnverifiedToolJet MCP is an implementation of this protocol on the ToolJet low-code platform, enabling coding agents to directly read the workspace structure, create data tables and queries, and generate interface components.
50%UnverifiedThe app uses three ToolJet groups, Support Sorter - Agents, Leads, and Viewers, to divide responsibilities: only Agents and admins can execute sorting, Jev, save, and note operations, Reset is restricted to admins, and Leads and Viewers are read-only; the rules are enforced at the query level, and blocked users receive a server-side rejection.
50%UnverifiedToolJet provides governance capabilities at the platform layer, supporting SAML, OIDC, or LDAP single sign-on with SCIM automated account provisioning, role-based access control refined down to apps, data sources, and each query, audit logs recording every login, edit, and approval decision, Docker or Kubernetes self-hosted air-gapped deployment, multi-user collaborative editing with version management and Git sync, and ToolJet AI can run within its own deployment environment.
50%UnverifiedIn an application built with ToolJet, a viewer calling a write query receives a 401 rejection, indicating that the server is indeed enforcing RBAC rules.
50%UnverifiedThe official ToolJet blog showcases a 'Support Ticket Sorter' app generated by a coding agent, using ToolJet MCP to build the app skeleton and TypeSafe's Jev to make the sorting judgments, with 100 synthetic tickets ultimately turned into a reviewable processing queue.
50%