bobbai00 opened a new pull request, #4417: URL: https://github.com/apache/texera/pull/4417
### What changes were proposed in this PR? Introduces the **Texera Agent** — an LLM-driven workflow copilot — in three layers: - **`agent-service/`** (new, Bun/TypeScript): ReAct-style agent that drives Texera workflows via a tool API (workflow, metadata, execution, code-op tools), with context assembly, result summarization, and Dagster / Hamilton Python sidecars. - **Backend**: new `SyncExecutionResource` for synchronous REST-driven runs used by the agent, a `PythonCodeValidator`, UDF descriptor updates for codegen, and deployment wiring (Helm templates for agent-service + LiteLLM, docker-compose, CI image build). - **Frontend**: workspace copilot UI (`agent-panel`, `agent-chat`, `agent-registration`, `agent-interaction`, `inline-code-panel`), a `TexeraCopilotManagerService`, and per-operator feedback entry point. Port-level result caching is intentionally **not** included in this PR. ### Any related issues, documentation, discussions? _To be filled in by the author._ ### How was this PR tested? Built locally and verified clean compilation on all three components: - `sbt "project WorkflowExecutionService" compile` - `cd agent-service && bun run typecheck` - `cd frontend && yarn build` (ng production) End-to-end agent flows were exercised manually against a running backend + agent-service. ### Was this PR authored or co-authored using generative AI tooling? Generated-by: Claude Code (Opus 4.7) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
