Open Issues Need Help
View All on GitHubAn AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: This GitHub issue identifies `json_default` in `llmchat_router.py` as dead code. The function is defined but never called within the codebase, appearing only in docstring examples. The suggested fix is to either remove the unused function or integrate it where JSON serialization is needed.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: The IBM Context Forge Gateway is rejecting 5 out of 12 critical Loki query tools during backend registration. This issue stems from an overly strict validation process that incorrectly flags backtick characters (`) in tool descriptions as "unsafe." As a result, essential log querying functionality is blocked because these tools are not exposed through the gateway.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: The `/tags/{tag_name}/entities` endpoint fails on PostgreSQL with a 500 error because the `get_entities_by_tag()` method in `tag_service.py` incorrectly uses the SQLite-specific `json_extract()` function. The fix requires replacing `func.json_extract()` with the cross-database `json_extract_text()` helper to properly handle JSON extraction on PostgreSQL.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: This is a feature request to add word wrap functionality to the CodeMirror editor instances within the `mcpgateway` admin panel UI. The goal is to improve readability for long lines of text or code, such as JSON content, by eliminating the need for horizontal scrolling.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: The task is to modify the Makefile to allow users to choose between using Docker or Podman for building and running the container image, ensuring consistent behavior and correct image name usage across both systems. The current Makefile likely has hardcoded references to either Docker or Podman, requiring a refactoring to use conditional logic based on user selection.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: This task requires enhancing the Admin UI's Metrics tab to include virtual servers in the metrics display and replace simple lists in the top performers section with detailed tables showing performance statistics (executions, response times, success rates, last execution time) for each entity type (tools, resources, prompts, gateways, servers). This involves backend modifications to fetch and format the detailed metrics and frontend changes to create responsive and visually appealing tables with appropriate styling and tab navigation.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: Implement file-based logging for the MCP Gateway, including log rotation, configurable log directory and file name, support for different log levels (console vs. file), and structured logging (JSON) for easier parsing and integration with log management systems. The implementation should also include log cleanup utilities to remove old log files and provide log statistics. Testing should verify all aspects of the new logging functionality.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: Update the MCP Gateway's version info page to clearly display the Redis cache configuration and connection status. This involves modifying the HTML template to show whether Redis is configured, connected, or not configured, and using appropriate styling (neutral, success, or error) for each state. The change should also clearly display the type of cache being used.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: The task requires fixing a console warning in the admin UI of the MCP Gateway by modifying the JavaScript code to handle cases where a loading element ('metrics-loading') might not exist in the DOM. This involves updating the `safeGetElement` function to optionally suppress warnings and modifying the metrics loading functions to use this updated function and add defensive checks. The solution should ensure no console warnings, correct loading states, and no visual regressions.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: Implement a Docker HEALTHCHECK instruction in the MCP Gateway's `Containerfile.lite` to ensure the container passes health checks promptly. This involves adding a Python-based health check command that verifies the application's health endpoint (`/health`) is responding correctly. The solution should handle potential startup delays and network connectivity issues, and be compatible with minimal container images. Testing should verify the health check's functionality during startup and under various conditions.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: The task requires extending the Streamable HTTP transport in the MCP Gateway to fully support virtual servers. This involves adding missing MCP method handlers for listing resources and prompts, updating the existing tool call handler to include server context, implementing robust error handling and validation for virtual server access, and ensuring consistent behavior with the existing SSE transport. The implementation should maintain backward compatibility and include comprehensive unit and integration tests.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: The task is to debug and fix a bug in the Model Context Protocol (MCP) Gateway's Admin UI. The bug causes all resource data, including metadata like name and URI, to be displayed within the 'content' field instead of being separated into their respective fields. The fix requires modifying the UI code to correctly handle resource data and prevent the unintended nesting of fields during editing.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: Debug and fix a bug in the Model Context Protocol (MCP) Gateway's Admin UI where edit screens for resources, prompts, and gateways are not populating fields correctly with existing data. This involves inspecting the UI code, identifying the data flow issue (likely a problem with fetching or binding data from the backend API to the frontend form), and implementing a fix to ensure that the edit screens correctly pre-fill the fields with the appropriate data from the database.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: Debug and fix a UI rendering issue in the MCP Gateway's admin panel where text boxes display a large empty space after line numbers when pasting content, and line numbers are unexpectedly deleted with backspace. The fix should ensure text aligns correctly to the left and line numbers are not removed unintentionally.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: Fix a bug in the MCP Gateway's admin UI where annotations in the tool edit screen are not editable and do not respect the dark mode setting. This involves updating the UI code to make the annotation fields editable and ensuring consistent styling across different modes.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: The task is to improve the consistency of error message display in the MCP Gateway's UI and API. This involves standardizing how invalid inputs (e.g., unsafe URLs, invalid names, malformed URIs) are handled and reported across different sections of the admin panel and API responses, aiming to eliminate internal server errors caused by invalid user input and replace them with clear, consistent warnings.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: This task requires adding REST API endpoints to an existing time server application that already uses the Model Context Protocol (MCP). The new REST endpoints should provide functionality for getting system time, converting time between time zones, listing time zones, getting timezone information, and performing batch conversions. The task also includes creating a testing suite for the new REST endpoints and updating the documentation to include REST API examples. The implementation should maintain existing MCP functionality and adhere to REST best practices.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: Improve the accessibility of the MCP Gateway Admin UI's dark theme by increasing the contrast and visibility of form input fields for tools. This involves identifying and updating CSS classes to ensure sufficient contrast between input fields and the dark background, testing across different browsers, and potentially implementing CSS variables for theme colors.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: The task requires updating the Makefile to correctly execute Alembic database migration commands. This involves adding the `-c $(ALEMBIC_CONFIG)` flag to all Alembic commands to specify the configuration file path, improving error handling and adding commands for checking migration status, creating new migrations, upgrading, downgrading, and resetting the database. Additionally, documentation in the README needs to be updated to reflect the changes.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: Improve the user experience of the MCP Gateway API by replacing raw technical error messages (like Pydantic validation errors and SQL exceptions) with user-friendly, concise messages. This involves creating a centralized error transformer to handle various error types and integrating it into the API's exception handling mechanism. The goal is to provide clear, actionable feedback to users while maintaining detailed logging for debugging purposes.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: The task requires fixing a broken 'Test Server Connectivity' feature in the Admin UI of an MCP Gateway. This involves debugging JavaScript code that handles form submission and API calls to test gateway connections, ensuring proper payload construction, error handling, and visual feedback. The fix should handle various transport types (HTTP, SSE, WebSocket) and provide clear success/failure messages with details. Improvements to the UI form and error handling are also suggested.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: This task requires extending an existing Python script to validate and automatically fix SPDX license headers in all Python source files. This involves creating new Makefile targets for validation and automated fixing, integrating the script into the CI/CD pipeline (GitHub Actions), and updating documentation. The goal is to ensure all files have consistent and correct SPDX headers for license compliance and accurate SBOM generation.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: Create a Python script to standardize headers in all project's Python files. The script should add or correct a header including the relative file path, copyright, license, and author information. The script should offer interactive and batch modes and integrate with the project's Makefile and potentially pre-commit hooks. Thorough testing and documentation are required.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: The task is to create documentation explaining how to use the `fast-time-server` MCP server via JSON-RPC commands using curl or stdio, covering health checks, version information, initialization, tool listing, and tool calls (including examples for `get_system_time` and `convert_time`). The documentation should also include instructions for using the server with Claude Desktop and troubleshooting common issues.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: This task requires implementing contextual hover help tooltips in the existing Admin UI of the MCP Gateway & Registry project. This involves adding tooltips to creation buttons (Add New Tool, Add New Server, etc.) and action buttons (Test, Activate/Deactivate, Delete) across various catalog panels. The tooltips should be accessible via keyboard navigation and adhere to IBM accessibility guidelines, including appropriate ARIA attributes and sufficient contrast in both light and dark modes. The tooltip text is specified in a provided copy map.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: The task is to implement a new feature in the MCP Gateway's Admin UI that allows administrators to test the connectivity and functionality of configured MCP servers by sending custom HTTP requests (GET, POST, PUT, DELETE) with configurable headers and body. The feature should provide feedback on the HTTP status code, response time, and response body, aiding in debugging server setup issues.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.
AI Summary: Implement configurable connection retry logic for database and Redis connections in the MCP Gateway. This involves adding new environment variables (`DB_MAX_RETRIES`, `DB_RETRY_INTERVAL_MS`, `REDIS_MAX_RETRIES`, `REDIS_RETRY_INTERVAL_MS`) to control the number of retries and retry intervals, updating the database and Redis connection logic to incorporate retries with exponential backoff, and updating documentation to reflect the changes.
An AI Gateway, registry, and proxy that sits in front of any MCP, A2A, or REST/gRPC APIs, exposing a unified endpoint with centralized discovery, guardrails and management. Optimizes Agent & Tool calling, and supports plugins.