Feature docker mcp #296
Open
Feature docker mcp #296
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
- Add dependency to enable MCP integration. - Implement to manage connections and aggregate tools from multiple MCP servers. - Create to handle communication with individual MCP servers via stdio. - Add a secure utility for executing whitelisted commands like Docker MCP Toolkit's CLI - Manage your MCP servers and clients. Usage: docker mcp [OPTIONS] Flags: -v, --version Print version information and quit Available Commands: catalog Manage MCP server catalogs client Manage MCP clients config Manage the configuration feature Manage experimental features gateway Manage the MCP Server gateway policy Manage secret policies secret Manage secrets server Manage servers tools Manage tools version Show the version information. - Define types for MCP server configuration, tools, and results.
…ation - Refactor to connect to servers and load tools on-demand rather than on initialization. - Introduce to wrap any MCP tool, making it compatible with the LangGraph agent framework. - Update to dynamically discover and load all available MCP tools at runtime. - Improve application startup performance by deferring MCP connections until a tool is needed.
- Refactor Docker command execution into a centralized utility with timeout support. - Fix MCP client reconnect logic to use proper exponential backoff with a max delay. - Make initialization asynchronous to prevent race conditions. - Harden JSON schema to Zod conversion for tool arguments. - Add a Mermaid sequence diagram documenting the MCP initialization and tool call flow.
- Update tool storage to support multiple tools with the same name from different servers. - Introduce a Docker Gateway mode for a unified tool catalog. - Allow disambiguating tool calls using serverName.toolName syntax. - Add retry logic to the server connection process for improved resilience. - Simplify the PR generation prompt by removing several instructional sections. - Move the Docker MCP Toolkit's CLI - Manage your MCP servers and clients. Usage: docker mcp [OPTIONS] Flags: -v, --version Print version information and quit Available Commands: catalog Manage MCP server catalogs client Manage MCP clients config Manage the configuration feature Manage experimental features gateway Manage the MCP Server gateway policy Manage secret policies secret Manage secrets server Manage servers tools Manage tools version Show the version information availability check into the MCP service.
- Introduce support for SSE (Server-Sent Events) transport to connect to remote MCP servers. - Implement automatic reconnect and retry logic for tool calls on connection failure. - Update MCP server configuration to include and options. - Enhance tool failure error messages to include a list of available tools. - Add to robustly handle different message formats from the agent.
- Add a new GET endpoint to fetch a single user. - Add dependency for JSON serialization. - Implement in the model. - Configure Jackson to not fail on empty beans.
This commit introduces a comprehensive overhaul of the agent's feedback mechanism and user interface, inspired by modern agentic systems like Claude. The previous implementation, which involved long waits followed by raw data dumps, is replaced with a real-time, transparent, and professional user experience. - Add a real-time activity feed to show the agent's current task (e.g., searching web, analyzing code). - Implement specialized UI cards for search results, code analysis, and errors, replacing raw markdown output. - Refactor the agent service to stream structured events (tool start/end, planning, summarizing) for detailed feedback. - Introduce a to transform raw tool outputs into clean summaries and structured data. - Enhance the MCP service for more robust Docker discovery, connection management, and asynchronous tool loading. - Add a detailed design and roadmap document for the new agent feedback system.
…activity streaming - Introduce detailed agent activity streaming for events like , , , providing real-time UI feedback. - Implement a human-in-the-loop (HITL) approval flow, requiring user consent only for file deletion. - Add robust safety mechanisms including iteration limits, tool call frequency limits, and timeouts to prevent infinite loops. - Introduce to prevent accidental VS Code closure while an agent is active. - Redesign the agent activity feed in the UI for a more compact and informative display. - Enhance the agent's system prompt to enforce tool usage after declaring intent. - Disable the chat input and display an approval prompt in the UI during agent execution.
- Implement a comprehensive, slide-in settings panel with sidebar navigation, replacing the old tab-based UI. - Add new settings sections for managing MCP servers, Custom Rules, Subagents, and more. - Create a reusable component library and for the new UI. - Introduce extension-side message handlers for managing MCP, rules, and subagents. - Define new configuration properties for custom rules, system prompts, and subagents in . - Add a detailed roadmap document for the settings UI overhaul.
- Integrate OpenAI, Qwen (DashScope), and GLM (Zhipu AI) as new model providers. - Add LLM classes and WebView providers for each new model, using an OpenAI-compatible interface. - Update configuration in to include settings for API keys and model names. - Extend the command handler to manage model creation, API calls, and chat history for the new providers. - Add new unit tests for the Qwen and GLM providers. - Update the UI dropdown to include the newly supported models.
Feature mcp contd
- Add as a new dependency - Implement method in the filesystem backend to read file content and metadata - Loosen agent type definitions in the factory and Gemini LLM for broader compatibility
feat(settings): introduce new sidebar-based settings panel
- Integrate providers for OpenAI, Qwen (Alibaba), and GLM (Zhipu). - Add new models to the selection dropdown in the settings UI. - Update FAQ documentation to include information on the new models. - Refactor the model settings component to use a centralized context.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.