mcpField
R
runtime

runtime

github.com/copilotkitAgent workflows38.3k2.7k installsUpdated today
Repository

About runtime

runtime is an index skill for developers configuring the server-side @copilotkit/runtime package. It directs readers to focused references for mounting endpoints, selecting agent runners, adding middleware, and connecting built-in or external agents.

Key features

  1. 01Maps tasks to runtime references
  2. 02Guides fetch-native endpoint setup
  3. 03Covers agent runner choices
  4. 04Documents middleware and request hooks
  5. 05Explains built-in and external agent wiring

Use cases

  1. 01Mounting a CopilotKit runtime handler
  2. 02Choosing persistence and runner behavior
  3. 03Adding authentication or request guards
  4. 04Connecting an external agent framework

MCP support

The runtime can be configured with MCP Apps middleware through its optional `mcpApps` setting. The input does not describe an MCP server interface.

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/copilotkit/copilotkit --skill runtime

After installing, type/runtimein Claude Code to confirm it loaded.

Bundled files

References and scripts installed along with SKILL.md.

references/
agent-runners-custom.md
agent-runners-in-memory.md
agent-runners-sqlite.md
agent-runners.md
built-in-agent-factory-modes.md
built-in-agent-helper-utilities.md
built-in-agent-model-identifiers.md
built-in-agent.md
intelligence-mode.md
middleware.md
server-side-tools.md
setup-endpoint.md
transcription.md
wiring-a2a.md
wiring-adk.md
wiring-ag2.md
wiring-agno.md
wiring-aws-strands.md
wiring-crewai-crews.md
wiring-crewai-flows.md
wiring-external-agents.md
wiring-langgraph.md
wiring-llamaindex.md
wiring-mastra.md
wiring-mcp-apps-middleware.md
wiring-ms-agent-framework.md
wiring-pydantic-ai.md

Details

From the project README

CopilotKit Runtime

@copilotkit/runtime is the server half of CopilotKit: it accepts AG-UI protocol requests, dispatches them to an AbstractAgent (built-in or external), runs the stream through an AgentRunner, and responds as Server-Sent Events.

This SKILL.md is the index. Read the reference under references/ that matches your task — do not try to absorb the whole package from this file.

Mental Model — the three dictionaries you hand to CopilotRuntime

new CopilotRuntime({
  agents, // Record<string, AbstractAgent>     — see wiring-external-agents or built-in-agent
  runner, // AgentRunner (optional)            — see agent-runners
  intelligence, // CopilotKitIntelligence (optional) — see intelligence-mode (auto-wires runner)
  mcpApps, // McpAppsConfig (optional)          — see wiring-mcp-apps-middleware
  a2ui, // A2UIConfig (optional)             — see packages/a2ui-renderer skill
  hooks, // { onRequest, onBeforeHandler }    — see middleware
  beforeRequestMiddleware,
  afterRequestMiddleware, // legacy — see middleware
  transcription, // TranscriptionService (optional)  — see transcription
});

You then mount it:

import { createCopilotRuntimeHandler } from "@copilotkit/runtime/v2";
const handler = createCopilotRuntimeHandler({
  runtime,
  basePath: "/api/copilotkit",
});
export default { fetch: handler };

When to load which reference

TaskReference
Mounting on any fetch-native server (Cloudflare Workers, Bun, Deno, Vercel Edge, Next.js App Router, React Router v7, TanStack Start) or delegating from Express/Nodereferences/setup-endpoint.md
Auth / logging / rate-limit / request-scoped guards via hooks.onRequest / hooks.onBeforeHandler (preferred) or legacy beforeRequestMiddleware / afterRequestMiddlewarereferences/middleware.md
Choosing between InMemoryAgentRunner, SqliteAgentRunner, or a custom subclass — including thread-locking semantics and the runner/Intelligence mutual exclusionreferences/agent-runners.md (+ -in-memory.md, -sqlite.md, -custom.md for backend-specific detail)
Enabling durable threads + realtime websocket via CopilotKit Intelligence (a managed service, not self-hostable)references/intelligence-mode.md
Voice transcription — implementing a TranscriptionService subclass for the /transcribe endpointreferences/transcription.md
Instantiating BuiltInAgent — Simple Mode (classic) or Factory Mode with TanStack AI (preferred AG-UI-compliant default), AI SDK, or custom factoryreferences/built-in-agent.md (+ -factory-modes.md, -helper-utilities.md, -model-identifiers.md)
Defining server-side tools via defineTool for BuiltInAgent.config.tools (Simple Mode only)references/server-side-tools.md
Wiring an external agent framework into CopilotRuntime({ agents })references/wiring-external-agents.md (index) + per-framework refs (wiring-mastra.md, wiring-langgraph.md, wiring-crewai-crews.md, wiring-crewai-flows.md, wiring-pydantic-ai.md, wiring-adk.md, wiring-llamaindex.md, wiring-agno.md, wiring-aws-strands.md, wiring-ms-agent-framework.md, wiring-ag2.md, wiring-a2a.md)
Wiring MCP Apps (runtime-level middleware, not an agent)references/wiring-mcp-apps-middleware.md

Invariants and gotchas (load-once, before any reference)

  • createCopilotRuntimeHandler is the canonical primitive. createCopilotExpressHandler / createCopilotHonoHandler exist but are avoid at all costs — delegate from Express/Hono routes to the fetch primitive instead.
  • Intelligence credentials are server-side. The CLI writes CPK_INTELLIGENCE_API_KEY into the runtime's environment; no client-side key is involved.
  • Intelligence mode auto-wires IntelligenceAgentRunner. Passing both runner and intelligence to CopilotRuntime is rejected at construction.
  • Intelligence mode targets the managed CopilotKit Intelligence service (api.cloud.copilotkit.ai) and is not self-hostable.
  • hooks.onRequest runs before beforeRequestMiddleware (hook-based middleware wins for Response short-circuits). beforeRequestMiddleware runs after hooks.onRequest (see fetch-handler.ts:136-147).
  • identifyUser (Intelligence) does not forward thrown Response objects — convert to 500. Gate auth rejection in hooks.onRequest, which does forward Responses.
  • agents__unsafe_dev_only and selfManagedAgents are dev-only aliases of each other; do not reach for them in production. Either signals that the SPA is in dev mode.

Reading order for a first-time reader

  1. setup-endpoint — the primitive.
  2. built-in-agent or pick one from wiring-external-agents — the agent.
  3. agent-runners — production persistence choice.
  4. Optional: middleware, intelligence-mode, server-side-tools, transcription.

Comments

Markdown supported