patterns
Design patterns for the Langroid multi-agent LLM framework. Covers agent configuration, tools, task control, and integrations.
What this skill does
# Langroid Patterns ## Instructions Below is an INDEX of design patterns organized by category. Each item describes WHAT you might want to implement, followed by a REFERENCE to a document with a complete code example. Scan this index to find patterns matching your needs, then consult the corresponding document. --- ## Agent & Task Basics 1. **Task Returns Tool Directly** Create a Langroid Agent equipped with a single Tool (a ToolMessage), and wrap it in a Task so that running the task returns that ToolMessage directly. Use this pattern when you want a simple LLM agent that returns a structured response. - Reference: `./task-return-tool.md` --- ## Tool Handlers 2. **Stateful Handler on Agent** Define a STATEFUL tool handler as a METHOD on the agent (not inside the ToolMessage). Use this pattern when: (a) the tool handler needs to execute external operations (API calls, database queries, file I/O), (b) you need to track state across retries (e.g., failure counter), (c) the handler needs access to agent-level resources (connections, configs), or (d) you want Langroid to automatically loop errors back to the LLM for self-correction. The method name must match the `request` field of the ToolMessage. Return a string for errors (LLM sees it and can retry), or DoneTool(content=result) to terminate successfully. - Reference: `./agent-tool-handler-with-state.md` 3. **Handler with Validation** Validate tool output against agent state before accepting it. Use this pattern when: (a) the LLM's tool output must preserve certain content from the input (e.g., placeholders, required fields), (b) you want automatic retry if validation fails, (c) you need to compare tool output against context the LLM received. Define a handler method on a custom agent class that stores the input context as state, validates the tool output, and returns an error string for retry or AgentDoneTool for success (note: use AgentDoneTool, NOT DoneTool). Use `done_sequences=["T[ToolName], A"]` so the handler runs before task termination. - Reference: `./agent-handler-validation-with-state.md` --- ## Task Control 4. **Terminate on Specific Tool** Terminate a Task only when a SPECIFIC tool is called. Use `TaskConfig(done_sequences=["T[ToolName]"])` to exit immediately when that tool is emitted, or `TaskConfig(done_sequences=["T[ToolName], A"])` to exit after the tool is emitted AND handled by the agent. Use this when an agent has multiple tools but you only want one specific tool to trigger task termination. - Reference: `./done-sequences-specific-tool.md` 5. **Batch Processing** Run the SAME task on MULTIPLE inputs concurrently using `run_batch_tasks()`. Use this pattern when: (a) you need to process many items with the same agent/task logic, (b) you want parallelism without manual asyncio/threading, (c) you need state isolation between items (each gets a cloned agent with fresh message history), (d) you want to avoid connection exhaustion from creating too many agents manually. Each item gets a cloned task+agent, runs independently, results collected in order. Supports batch_size for concurrency limiting. - Reference: `./run-batch-tasks.md` --- ## Integration & Output 6. **MCP Tools Integration** Enable a Langroid agent to use MCP (Model Context Protocol) tools from an external MCP server like Claude Code. Use this pattern when: (a) you want your agent to use file editing tools (Read, Edit, Write) from Claude Code, (b) you need to connect to any MCP server via stdio transport, (c) you want to enable ALL tools from an MCP server or just SPECIFIC tools selectively, (d) you want to customize/post-process MCP tool results before returning to the LLM. Uses `@mcp_tool` decorator for specific tools or `get_tools_async()` for all tools. - Reference: `./mcp-tool-integration.md` 7. **Quiet Mode** Suppress verbose Langroid agent output (streaming, tool JSON, intermediate messages) while showing your own custom progress messages. Use this pattern when: (a) you want clean CLI output showing only milestone events, (b) you're running a multi-step workflow and want to show progress without agent noise, (c) you need thread-safe output control. Use `quiet_mode()` context manager to wrap agent task.run() calls, then print your own messages outside the context. - Reference: `./quiet-mode.md`
Related in Design
contribute
IncludedLocal-only OSS contribution command center. Auto-refreshes the user's in-flight PR and issue state on invoke so conversations start with full context — no need to brief Claude on what's in flight. Helps the user find issues to contribute to on GitHub, builds per-repo dossiers of what each upstream expects (CLA, DCO, branch convention, AI policy, draft-first, review bots, issue templates), runs deterministic gates before any external action so AI-assisted contributions don't reach maintainers as slop. State is markdown-only: candidate files at ~/.contribute-system/candidates/, repo dossiers at ~/.contribute-system/research/, append-only event log at ~/.contribute-system/log.jsonl. No database, no cloud calls. Use when the user asks about their PRs / issues / contributions, wants to find new work to take on, claim an issue, build/refresh a repo's dossier, or draft a Design Issue or PR. Trigger with "/contribute", "what's my PR status", "find a contribution", "claim issue X", "draft a Design Issue for Y", "refresh dossier for Z".
architectural-analysis
IncludedUser-triggered deep architectural analysis of a codebase or scoped subtree across eight modes — information architecture, data flow, integration points, UI surfaces, interaction patterns, data model, control flow, and failure modes. This skill should be used when the user asks to "diagram this codebase," "map the architecture," "show the data flow," "give me an ERD," "trace control flow," "find the integration points," "verify the layout pattern," "audit the UX architecture," or any similar request whose primary deliverable is mermaid diagrams plus cited reports under docs/architecture/. Dispatches haiku/sonnet sub-agents in parallel for per-mode exploration, then verifies every citation mechanically before any node lands in a diagram. Not for one-off prose explanations of code (use code-explanation) or for high-level system design from scratch (use system-design).
mcp
IncludedModel Context Protocol (MCP) server development and tool management. Languages: Python, TypeScript. Capabilities: build MCP servers, integrate external APIs, discover/execute MCP tools, manage multi-server configs, design agent-centric tools. Actions: create, build, integrate, discover, execute, configure MCP servers/tools. Keywords: MCP, Model Context Protocol, MCP server, MCP tool, stdio transport, SSE transport, tool discovery, resource provider, prompt template, external API integration, Gemini CLI MCP, Claude MCP, agent tools, tool execution, server config. Use when: building MCP servers, integrating external APIs as MCP tools, discovering available MCP tools, executing MCP capabilities, configuring multi-server setups, designing tools for AI agents.
react-native-skia
IncludedDesign, build, debug, and optimise high-polish animated graphics in React Native or Expo using @shopify/react-native-skia, Reanimated, and Gesture Handler. Use when the user wants canvas-driven UI, shaders, paths, rich text, image filters, sprite fields, Skottie, video frames, snapshots, web CanvasKit setup, or performance tuning for custom motion-heavy elements such as loaders, hero art, cards, charts, progress indicators, particle systems, or gesture-driven surfaces. Also use when the user asks for fluid, glow, glass, blob, parallax, 60fps/120fps, or GPU-friendly animated effects in React Native, even if they do not explicitly say "Skia". Do not use for ordinary form/layout work with standard views.
plaid
IncludedProduct Led AI Development — guides founders from idea to launched product. Six capabilities: Idea (discover a product idea), Validate (pressure-test the idea against fatal flaws, problem reality, competition, and 2-week MVP feasibility), Plan (vision intake + document generation), Design (translate image references into a design.md spec), Launch (go-to-market strategy), and Build (roadmap execution). Use when someone says "PLAID", "plaid idea", "help me find an idea", "product idea", "idea from my business", "idea from my expertise", "plaid validate", "validate my idea", "pressure-test", "is this idea good", "find fatal flaws", "validate the problem", "plan a product", "define my vision", "generate a PRD", "product strategy", "plaid design", "design from image", "translate image to design", "create design.md", "extract design tokens", "plaid launch", "go-to-market", "launch plan", "GTM strategy", "launch playbook", "plaid build", "build the app", "start building", or "execute the roadmap".
nextjs-framer-motion-animations
IncludedAdds production-safe Motion for React or Framer Motion animations to Next.js apps, including reveal, hover and tap micro-interactions, whileInView, stagger, AnimatePresence, layout and layoutId transitions, reorder, scroll-linked UI, and lightweight route-content transitions. Use when the user asks to add, refactor, or debug Motion or Framer Motion in App Router or Pages Router codebases, especially around server/client boundaries, reduced motion, LazyMotion, bundle size, hydration, or route transitions. Avoid for GSAP-style timelines, WebGL or 3D scenes, heavy scroll storytelling, or CSS-only effects unless Motion is explicitly requested.