Scala good first issues

AI Summary: This issue proposes consolidating the Apache Kyuubi documentation from a hybrid reStructuredText (.rst) and Markdown (.md) format to a single approach. The goal is to simplify the contribution process, improve cross-referencing, and streamline the build pipeline by eliminating the need for multiple Sphinx extensions.

AI Summary: The scalafix CLI tool silently ignores repeated --classpath arguments if they don't contain SemanticDB information, leading to unexpected errors when trying to analyze code that depends on libraries without SemanticDB. The user expects that non-SemanticDB JARs can be included in the classpath for dependency resolution without causing errors.

AI Summary: This issue proposes deleting the `package.json` file, assuming there are no longer any npm dependencies. The author suggests trying this and observing any resulting breakage to confirm the assumption.

AI Summary: The `lower_enabled.sql` and `upper_enabled.sql` test files are not correctly testing the native case-conversion path as intended. They are configured to use a no-op setting, causing them to always fall back to the codegen dispatcher instead of the native implementation. This issue appears to be a leftover from a previous configuration change and is not causing any visible failures due to the test data.

AI Summary: The task requires fixing a bug in the Delta Lake Kernel component. The bug causes an exception when a staged commit file is renamed during pagination, specifically when a file's name changes from a format including a UUID to a simpler format (e.g., 9.uuid.json to 9.json). The solution involves modifying the log segment validation to correctly handle this type of file renaming without throwing an exception.

AI Summary: This issue describes a bug where `named_struct` with duplicate field names causes the entire projection to fall back to Spark. The fix involves mixing `CodegenDispatchFallback` into `CometCreateNamedStruct` to handle this scenario, as Spark and Arrow allow duplicate field names in structs.

AI Summary: This issue proposes to enable Comet's `length`, `bit_length`, and `octet_length` functions to handle `BinaryType` input. Currently, these functions fall back to Spark for binary input, which is inefficient. The fix involves mixing `CodegenDispatchFallback` into these functions, allowing them to be handled natively by Comet.

AI Summary: The `translate` string function in Comet currently falls back to Spark by default due to an unconditional `Incompatible` status, even though other string functions use a codegen dispatcher. This issue proposes mixing `CodegenDispatchFallback` into `CometStringTranslate` to enable default codegen support, similar to other string functions, with the grapheme/code-point difference becoming an opt-in caveat.

AI Summary: The `abs` function applied to interval types in Comet currently causes a fallback to Spark's implementation, even though Comet's dispatcher explicitly supports these interval types. This issue proposes mixing `CodegenDispatchFallback` into `CometAbs` to address this, leveraging Spark's built-in ANSI overflow behavior for intervals.

AI Summary: This issue highlights a potential security concern in multi-tenant Kyuubi clusters where session configuration restrictions are not set by default. The author suggests improving documentation to guide administrators on configuring `kyuubi.session.conf.restrict.list` or `kyuubi.session.conf.ignore.list` to prevent users from overwriting session configurations, rather than changing default values.

AI Summary: This issue proposes adding integration tests for the workspace's containerized code execution feature. The tests will verify the end-to-end flow from an agent receiving a task, calling a workspace tool, executing code in Docker, and returning the result. Specific scenarios to be tested include basic execution, timeout enforcement, sandbox isolation, multi-language support, error propagation, and container cleanup.

AI Summary: This issue proposes adding integration tests for the Model Context Protocol (MCP) server connections. The goal is to verify end-to-end functionality, including connecting to an MCP server, listing tools, and invoking them through an agent. The preferred approach involves an embedded test MCP server for CI reliability, with an optional real server integration and mock-based tests also suggested.

AI Summary: This issue proposes adding integration tests for the agent's memory subsystem, specifically focusing on the `VectorMemoryStore` which uses embedding-backed storage. The tests will verify that memories are correctly recorded, retrieved, injected into agent conversations across multiple turns, isolated by user, and persist across different agent instances, requiring pgvector and Ollama to be available.

AI Summary: This issue proposes adding a comprehensive integration test for the RAG pipeline. The test will verify the end-to-end functionality of indexing documents, embedding them with Ollama, storing them in pgvector, performing semantic and hybrid searches, and finally using the retrieved context for agent-based completions. It aims to cover real-world usage scenarios that are currently untested.

AI Summary: This issue requests the creation of a new Gradle sample project for Java developers to demonstrate how to use the llm4s library. The project will include a `build.gradle.kts` file, a sample Java class calling the llm4s API, and a README with instructions on how to run it. This aims to lower the adoption barrier for Java shops.

AI Summary: This issue requests the creation of a new Scala sample file demonstrating the `CompositeGuardrail` feature. The sample should showcase how to chain multiple guardrails using `all()`, `any()`, and `sequence()` combinators, illustrating different logical compositions and their behavior with various test inputs. No LLM calls are required, focusing solely on rule-based guardrails.

AI Summary: This issue requests the creation of a new Scala sample demonstrating the multi-provider support of the llm4s library. The sample should run the same prompt against various configured LLM providers (OpenAI, Anthropic, Gemini, etc.) and display their results side-by-side, including response time and token count. The goal is to showcase the library's unified API for interacting with different LLM services.

AI Summary: This issue requires the creation of a comprehensive sample application and accompanying documentation to demonstrate the end-to-end functionality of cost tracking features. The sample should showcase per-request, agent-level, and session-level cost aggregation, as well as custom model pricing registration and integration with Prometheus metrics.

AI Summary: This issue proposes adding support for several major cloud provider image generation APIs to the llm4s library. The goal is to integrate services like Google Vertex AI (Imagen), AWS Bedrock, Stability AI's direct API, and Fal AI, following an established implementation pattern for new providers. Each provider integration is considered a separate, low-complexity task.

AI Summary: This issue proposes making HTTP timeouts configurable for all LLM provider clients, as they are currently hardcoded. The task involves adding a `timeoutMs` parameter to provider configurations, integrating it into HTTP client construction for each provider, and updating the main configuration to allow setting this value via an environment variable. A test should be added to verify that a short timeout causes immediate failure.

AI Summary: This issue requests the addition of unit tests for the `Schema` and `SchemaDefinition` classes in the `llm4s` project. The tests should verify that these classes correctly generate JSON Schema representations for various data types and configurations, ensuring that tool parameter schemas sent to LLM providers are accurate. A specific focus is placed on testing different schema types, handling of required/optional fields, and a round-trip verification with `ToolFunction`.

AI Summary: This issue requests the addition of comprehensive unit tests for the `SafeParameterExtractor` Scala class. The tests should cover various edge cases for extracting string, integer, and boolean parameters, as well as handling optional parameters and nested object access. The goal is to ensure robust parameter parsing for LLM tool calls and prevent production runtime errors.

AI Summary: This issue requests the creation of a new documentation guide for the llm4s project. The guide will explain how to work with `Result[A]` (which is an `Either[LLMError, A]`) and `LLMError` types in practice, targeting developers unfamiliar with Scala's `Either`. It will include explanations, code examples for basic patterns, for-comprehensions, error types, specific error handling, converting to exceptions, and testing, along with instructions on how to get started and acceptance criteria.