**Root Cause**
Many-to-one relation filters were being exposed under the relation name
(e.g., `company`) instead of the corresponding foreign-key attribute
(e.g., `companyId`).
**Change**
- Detect many-to-one relation metadata and remap those filter keys to
`<fieldName>Id`.
- Removed some unused code unrelated to this fix.
## Description
Fixes#15348
The Vercel AI SDK serializes tool schemas with a `jsonSchema` wrapper
(`{ jsonSchema: {...} }`), but the Model Context Protocol specification
expects the schema directly (`{ type: 'object', properties: {...} }`).
This was causing MCP clients like Claude Desktop to silently reject all
tool definitions as invalid, leaving users with no available tools.
## Changes
Modified `handleToolsListing()` in `mcp.service.ts` to unwrap the
`jsonSchema` key before returning tools to MCP clients.
## Before
```json
{
"name": "create_company",
"description": "...",
"inputSchema": {
"jsonSchema": {
"type": "object",
"properties": {...}
}
}
}
```
## After
```json
{
"name": "create_company",
"description": "...",
"inputSchema": {
"type": "object",
"properties": {...}
}
}
```
## Testing
Tested with:
- Direct curl request to verify schema format
- Claude Desktop client via mcp-proxy
- Verified tools now appear correctly in MCP clients
## References
- [MCP Tools
Specification](https://modelcontextprotocol.io/specification/2025-06-18/server/tools)
- Related discussion in #12953
This PR implements Sentry's AI agent monitoring by:
- Configuring vercelAIIntegration with recordInputs and recordOutputs
options
- Adding sendDefaultPii to Sentry.init() for better debugging
- Creating a shared AI_TELEMETRY_CONFIG constant to DRY up the telemetry
configuration
- Adding experimental_telemetry to all AI SDK calls (generateText,
generateObject, streamText)
All AI operations are now fully monitored in Sentry with complete
input/output recording for debugging and performance analysis.
Note: Currently on Sentry v9.26.0, which is compatible with this
implementation. No breaking changes from the v9-to-v10 migration guide
were found in the codebase.
Adds intelligent routing system that automatically selects the best
agent for user queries based on conversation context.
### Changes:
- Added `routerModel` column to workspace table for configurable router
LLM selection
- Implemented `RouterService` with conversation history analysis and
agent matching logic
- Created router settings UI in AI Settings page with model dropdown
- Removed agent-specific thread associations - threads are now
agent-agnostic
- Added real-time routing status notification in chat UI with shimmer
effect
- Removed automatic default assistant agent creation
- Renamed GraphQL operations from agent-specific to generic (e.g.,
`agentChatThreads` → `chatThreads`)
---------
Co-authored-by: Félix Malfait <felix.malfait@gmail.com>
Co-authored-by: Félix Malfait <felix@twenty.com>
Implements permission intersection (AND logic) to prevent permission
escalation when agents act on behalf of users.
### Changes:
- **Permission Intersection**: Operations requiring both user AND agent
permissions
- **RoleContext Type**: Unified type supporting single `roleId` or
multiple `roleIds` for intersection
- **CRUD Services**: Updated to accept `roleContext` for granular
permission control
- **Agent Integration**: Chat agents now use user + agent role
intersection for all operations
- **ORM Layer**: Enhanced `getRepository` to support multi-role
permission checks
### Related:
- Part 2 of ["Acting on behalf of user" concept
PR](https://github.com/twentyhq/twenty/pull/15103)
[Closes#1661](https://github.com/twentyhq/core-team-issues/issues/1661)
---------
Co-authored-by: Félix Malfait <felix.malfait@gmail.com>
## Summary
**Step 1 of 2:** Implements the "acting on behalf of user" concept for
workflows and agents to prevent permission escalation and maintain
proper audit trails.
## Problem
Previously, workflows and agents would bypass permissions regardless of
who initiated them, allowing users to escalate their privileges by
triggering workflows that performed actions they couldn't do directly.
## Solution
### For Workflows
Introduced `WorkflowExecutionContext` service that determines execution
mode:
- **Manual triggers/test button**: Uses user's roleId for permissions,
user's identity for `createdBy`
- **Automated triggers** (cron, database events, webhooks): Bypasses
permissions, uses workflow identity
### For Agents
**In Chat:**
- Always act on behalf of the user
- Use user's roleId for permission checks
- Use user's identity for `createdBy`
# Step 1 vs Step 2
### ✅ Step 1 (This PR): Acting on Behalf Concept
- Introduced `isActingOnBehalfOfUser` boolean concept
- Single roleId used for permission checks (user's OR system bypass)
- `createdBy` field properly attributes actions to initiator
- Prevents permission escalation in user-initiated flows
### 🔜 Step 2 (Future): Multi-Role Permission Support
- Support role intersection: `{ intersection: ['roleA', 'roleB'] }`
- Support role union: `{ union: ['roleA', 'roleB', 'roleC'] }`
- Enable user+agent collaboration scenarios
- Update `WorkspaceEntityManager` and `WorkspaceDatasource` to handle
multiple roleIds
---------
Co-authored-by: Félix Malfait <felix.malfait@gmail.com>
## TODO:
- [x] display "yearly" or "monthly" wording everywhere it's needed
- [ ] Add button with "downgrade" or "upgrade" to save the change of
credits price + modal to validate
- [x] Add renewal date
- [ ] Implement
https://docs.stripe.com/billing/subscriptions/subscription-schedules for
`switchFromYearlyToMonthly` and decrease number of credits
…PC validation
- Introduced full integration test suite for MCP controller, testing
`POST /mcp` with valid and invalid payloads.
- Added `@IsDefined` validation to ensure the `method` field is required
in JSON-RPC requests.
- Applied `RestApiExceptionFilter` to MCP controller for consistent
error handling.
- Enhanced validation pipe in MCP controller to whitelist and reject
non-whitelisted properties.
- Consolidated exception filters in SSOAuthController.
…s, and resources
- Removed unused `ObjectMetadataStandardIdToIdMap`.
- Updated Vite config with additional allowed hosts.
- Improved MCP service to handle `ping` method and lists for tools,
prompts, and resources.
- Refactored utility function `isFieldMetadataEntityOfType` for improved
type handling.
- Expanded MCP metadata service to include tools, prompts, and resources
support.
- Renamed `WorkflowActionAdapter` to `ToolExecutorWorkflowAction`
- Renamed `settingPermission` table to `permissionFlag` and `setting`
column to `flag`
- Decoupled the send email logic from workflows to tools
- Add new `Tools Permission` section in FE
---------
Co-authored-by: Félix Malfait <felix.malfait@gmail.com>
Co-authored-by: Félix Malfait <felix@twenty.com>
- Added a generic HTTP request tool, allowing agents and workflows to
make HTTP requests to external APIs with configurable method, headers,
and body.
- Decoupled HTTP request workflow nodes from workflow-specific types and
factories, introducing a generic tool interface.
- Updated agent system prompts to include explicit guidance for the HTTP
request tool, including when and how to use it, and how to communicate
limitations.
### Demo
https://github.com/user-attachments/assets/129bc445-a277-4a19-95ab-09f890f8f051
## Context
Add an eventEmitter instance to twenty datasources so we can emit DB
events.
Add input and output formatting to twenty orm (formatData, formatResult)
Those 2 elements simplified existing logic when we interact with the
ORM, input will be formatted by the ORM so we can directly use
field-like structure instead of column-like. The output will be
formatted, for builder queries it will be in `result.generatedMaps`
where `result.raw` preserves the previous column-like structure.
Important change: We now have an authContext that we can pass when we
get a repository, this will be used for the different events emitted in
the ORM. We also removed the caching for repositories as it was not
scaling well and not necessary imho
Note: An upcoming PR should handle the onDelete: cascade behavior where
we send DESTROY events in cascade when there is an onDelete: CASCADE on
the FK.
---------
Co-authored-by: Charles Bochet <charles@twenty.com>
Instead of initializing model at start time we do it at run time to be
able to swap model provider more easily.
Also introduce a third driver for openai-compatible providers, which
among other allows for local models with Ollama
martmull