### Background
We need a way to set an execution quota per user for each block execution.
### Changes 🏗️
* Introduced a `UserBlockCredit`, a transaction table tracking the block usage along with it cost/quota.
* The tracking is toggled by `ENABLE_CREDIT` config, default = false.
* Introduced `BLOCK_COSTS` | `GET /blocks/costs` as a source of information for the cost on each block depending on the input configuration.
Improvements:
* Refactor logging in manager.py to always print a prefix and pass the metadata.
* Make executionStatus on AgentNodeExecution prisma enum. And add executionStatus on AgentGraphExecution.
* Use executionStatus from AgentGraphExecution to improve waiting logic on test_manager.py.
- feat(builder): Add "Stop Run" buttons to monitor and builder
- Implement additional state management in `useAgentGraph` hook
- Add "stop" request mechanism
- Implement execution status tracking using WebSockets
- Add `isSaving`, `isRunning`, `isStopping` outputs
- Add `requestStopRun` method
- Rename `requestSaveRun` to `requestSaveAndRun` for clarity
- Add needed functionality for the above to `AutoGPTServerAPI` client
- Add `stopGraphExecution` method
- Add support for multiple handlers per WebSocket method
- Fix parsing of timestamps in `execution_event` WebSocket messages
- Add `IconSquare` from Lucide to `@/components/ui/icons`
- feat(server): Add `POST /graphs/{graph_id}/executions/{graph_exec_id}/stop` route
- Add `stop_graph_run` method to `AgentServer`
- feat(server): Add `cancel_execution` method to `ExecutionManager`
- Replace node executor `ProcessPoolExecutor` by `multiprocessing.Pool` (which has a `terminate()` method)
- Remove now unnecessary `Executor.wait_future(..)` method
- Add `get_graph_execution(..)` in `.data.execution`
- fix(server): Reduce number of node executors to 5 per graph executor
This is necessary because `multiprocessing.Pool` spawns its workers on init, instead of based on demand like `ProcessPoolExecutor` does
- dx(server): Improve debug logging in `ExecutionManager`
- ci(server): Add debug logging mode to CI Pytest step
### Other improvements
Server:
- Improve output type of `ExecutionManager.add_execution(..)`
- Renamed a few things in `.server.rest_api` for consistency
Front end:
- Improved typing in `AutoGPTServerAPI` client
### Background
When multiple executors are executing the same node within the same graph execution, two node executions can read the same queue of input and read the same value—making the data that is supposed to be consumed once, consumed by two executions. The lack of lock & concurrency support for parallel execution within a single graph causes this issue.
Node concurrency also introduces poor UX in the current frontend implementation, when two nodes are executed in parallel, the current UI will not display its parallel execution update, but instead, it shows the updates that override each other. Until the execution observability is improved on the builder UI, this capability will be limited.
### Changes 🏗️
The scope of this change is to solve this issue by:
* Decouple Graph execution & Node execution, each has its own configured process pool.
* Make sure there is only 1 execution per node (we still allow parallel executions on different nodes) in a graph.
* Fixed concurrency issue by adding distributed lock API on agent_server.
* Few cleanups:
- Add more logging with geid & neid prefix on graph/node executions
- Moved execution status update to agent-server for a single source of status update (required by conn-manager/web-socket)
- Configured node parallelism to 10 & graph parallelism to 10 by default, so in the very rare worst-case, there can be 100 node executions.
- Re-use server resource for each integration test run
