Vellum-ai

- Support WorkflowEventDisplayContext for BaseWorkflowDisplay

- Workflow Streaming View now supports expand_meta parameters cost and model name

- Bump the TypeScript generator version to include timeout fixes

- Support Chat History and other List of pydantic output types for templating nodes
- Support Templating node outputs for pydantic models
- Support versions of Jinja > 3.1.2

- Implement `--strict` on vellum push CLI
- Serialize input variable name with override
- Unblock pushes of generic nodes
- Codegen fails container serialization
- Serialize node attributes
- Serialize node outputs
- [Codegen][Part 2] Codegen triggers of generic nodes

- Adds support for serializing all reference types
- Adds support for serializing unary and ternary expressions
- Introduces a new `vellum ping` command to the CLI for active Workspace and Organization information
- Introduces new types for Search Node metadata filtering
- Type checking for Accessor Expression now accepts ints
- The `prompt_inputs` attribute on Prompt Nodes is now optional
- `MapNode.concurrency` attribute is now renamed to `MapNode.max_concurrency`
- `MapNode.max_concurrency` is now supported as an adornment parameter
- Fixed a bug with Node Output referencing of list types in future versions of Python
- Introduces a new `max_concurrency` parameter to define max number of threads in Workflow runs
- Adds support for serialization of AND and OR expressions in Node Ports
- Map Node now supports streaming events at the initiation and fulfillment of each subworkflow iteration
- Adds a new `strict` request body option to the /workflows/push endpoint
- Silences a warning when referencing a descriptor within a pydantic model used as a Node Attribute