@m4ike1/ion-agent-core

Stateful agent runtime with tool execution and event streaming. Built on @m4ike1/ion-ai.

The Agent class owns the transcript, runs the prompt → LLM → tool loop, and emits lifecycle events for UI updates. For details see docs/how-to.md (recipes) and docs/api-reference.md (API dictionary). The durable conversation/session harness in this package is specified separately in docs/harness.md.

Installation

npm install @m4ike1/ion-agent-core

Entrypoints

import { Agent } from "@m4ike1/ion-agent-core"; // core runtime (Agent, agentLoop, streamProxy, types)
import { NodeExecutionEnv } from "@m4ike1/ion-agent-core/node"; // Node execution env + everything above
import type { ... } from "@m4ike1/ion-agent-core/harness/session"; // session layer
import type { ... } from "@m4ike1/ion-agent-core/harness/session/testing"; // session test helpers
import { reduceLaneSnapshot } from "@m4ike1/ion-agent-core/harness/runtime/reducer";
import type { ... } from "@m4ike1/ion-agent-core/harness/context";
import { NodeExecutionEnv } from "@m4ike1/ion-agent-core/harness/env/nodejs";

SQLite session backends

The SQLite session backend and the node:sqlite adapter live in a separate package, @m4ike1/ion-session-backend-sqlite-node, so the core package does not pull in runtime builtins or native SQLite dependencies by default. The backend accepts a runtime-specific SQLite factory, allowing other session backends to ship as their own packages in the future.

Quick Start

import { Agent } from "@m4ike1/ion-agent-core";
import { createModels } from "@m4ike1/ion-ai";
import { anthropicProvider } from "@m4ike1/ion-ai/providers/anthropic";

const models = createModels();
models.setProvider(anthropicProvider());
const model = models.getModel("anthropic", "claude-sonnet-4-6");
if (!model) throw new Error("Model not found");

const agent = new Agent({
  initialState: {
    systemPrompt: "You are a helpful assistant.",
    model,
  },
  streamFn: models.streamSimple.bind(models),
});

agent.subscribe((event) => {
  if (event.type === "message_update" && event.assistantMessageEvent.type === "text_delta") {
    // Stream just the new text chunk
    process.stdout.write(event.assistantMessageEvent.delta);
  }
});

await agent.prompt("Hello!");

streamFn defaults to the function installed via setDefaultStreamFn() when omitted. Models.streamSimple satisfies the StreamFn contract: never throw for request/model/runtime failures; encode failures in the returned stream as a final AssistantMessage with stopReason "error" / "aborted".

Experimental facet services

Transport-neutral facet-service primitives live in @m4ike1/chord. The agent core does not export the service runtime.

Core Concepts

AgentMessage vs LLM Message

The agent works with AgentMessage, a flexible type that can include:

• Standard LLM messages (user, assistant, toolResult)
• Custom app-specific message types via declaration merging

LLMs only understand user, assistant, and toolResult. The convertToLlm function bridges this gap by filtering and transforming messages before each LLM call. The default convertToLlm passes through only user / assistant / toolResult.

Message Flow

AgentMessage[] → transformContext() → AgentMessage[] → convertToLlm() → Message[] → LLM
                    (optional)                           (required)

1. transformContext: Prune old messages, inject external context
2. convertToLlm: Filter out UI-only messages, convert custom types to LLM format

Both hooks must not throw or reject; return a safe fallback instead. A throwing convertToLlm, shouldStopAfterTurn, or prepareNextTurn interrupts the low-level loop without producing a normal event sequence.

Event Flow

The agent emits events for UI updates. Understanding the event sequence helps build responsive interfaces.

prompt() Event Sequence

When you call prompt("Hello"):

prompt("Hello")
├─ agent_start
├─ turn_start
├─ message_start   { message: userMessage }      // Your prompt
├─ message_end     { message: userMessage }
├─ message_start   { message: assistantMessage } // LLM starts responding
├─ message_update  { message: partial... }       // Streaming chunks
├─ message_update  { message: partial... }
├─ message_end     { message: assistantMessage } // Complete response
├─ turn_end        { message, toolResults: [] }
└─ agent_end       { messages: [...] }

With Tool Calls

If the assistant calls tools, the loop continues:

prompt("Read config.json")
├─ agent_start
├─ turn_start
├─ message_start/end  { userMessage }
├─ message_start      { assistantMessage with toolCall }
├─ message_update...
├─ message_end        { assistantMessage }
├─ tool_execution_start  { toolCallId, toolName, args }
├─ tool_execution_update { partialResult }           // If tool streams
├─ tool_execution_end    { toolCallId, result }
├─ message_start/end  { toolResultMessage }
├─ turn_end           { message, toolResults: [toolResult] }
│
├─ turn_start                                        // Next turn
├─ message_start      { assistantMessage }           // LLM responds to tool result
├─ message_update...
├─ message_end
├─ turn_end
└─ agent_end

Tool execution mode is configurable:

• parallel (default): preflight tool calls sequentially, execute allowed tools concurrently, emit tool_execution_end as soon as each tool is finalized, then emit toolResult messages and turn_end.toolResults in assistant source order
• sequential: execute tool calls one by one, matching the historical behavior

In parallel mode, tool completion events follow tool completion order, but persisted toolResult messages still follow assistant source order.

The mode can be set globally via toolExecution in the agent config, or per-tool via executionMode on AgentTool. If any tool call in a batch targets a tool with executionMode: "sequential", the entire batch executes sequentially regardless of the global setting.

The beforeToolCall hook runs after tool_execution_start and validated argument parsing. It can block execution and attach terminate: true to the blocked result. The afterToolCall hook runs after tool execution finishes and before tool_execution_end and final tool result message events are emitted.

Tools, blocked beforeToolCall results, and afterToolCall overrides can return terminate: true to hint that the automatic follow-up LLM call should be skipped. The loop only stops early when every finalized tool result in that batch sets terminate: true. Mixed batches continue normally.

When you use the Agent class, assistant message_end processing is treated as a barrier before tool preflight begins. That means beforeToolCall sees agent state that already includes the assistant message that requested the tool call.

Truncated responses

If the assistant message ends with stopReason: "length" (output token limit hit), no tool call in that message is executed: streamed tool-call arguments may be silently truncated, so each call is reported as an error result and the model is asked to re-issue the calls with complete arguments.

continue() Event Sequence

continue() resumes from existing context without adding a new message. Use it for retries after errors.

// After an error, retry from current state
await agent.continue();

The last message in context must be user or toolResult (not assistant). If the last message is assistant but a steering or follow-up message is queued, continue() runs the queued message instead of throwing.

Event Types

| Event | Description | |-------|-------------| | agent_start | Agent begins processing | | agent_end | Final event for the run. Awaited subscribers for this event still count toward settlement | | turn_start | New turn begins (one LLM call + tool executions) | | turn_end | Turn completes with assistant message and tool results | | message_start | Any message begins (user, assistant, toolResult) | | message_update | Assistant only. Includes assistantMessageEvent with delta | | message_end | Message completes | | tool_execution_start | Tool begins | | tool_execution_update | Tool streams progress | | tool_execution_end | Tool completes |

Agent.subscribe() listeners are awaited in registration order and receive the active run's AbortSignal. agent_end means no more loop events will be emitted, but await agent.waitForIdle() and await agent.prompt(...) only settle after awaited agent_end listeners finish. Raw agentLoop() / agentLoopContinue() streams are observational only: they preserve event order but do not wait for your async event handling to settle before later producer phases continue.

Agent Options

See docs/api-reference.md for the full field dictionary. Common setup:

const agent = new Agent({
  // Initial state
  initialState: {
    systemPrompt: "You are a helpful assistant.",
    model,
    thinkingLevel: "medium", // "off" | "minimal" | "low" | "medium" | "high" | "xhigh" | "max"
    tools: [readFileTool],
    messages: [],
  },

  // Convert AgentMessage[] to LLM Message[] (required for custom message types)
  convertToLlm: (messages) => messages.filter(...),

  // Transform context before convertToLlm (for pruning, compaction)
  transformContext: async (messages, signal) => pruneOldMessages(messages),

  // Steering mode: "one-at-a-time" (default) or "all"
  steeringMode: "one-at-a-time",

  // Follow-up mode: "one-at-a-time" (default) or "all"
  followUpMode: "one-at-a-time",

  // Stream function (falls back to setDefaultStreamFn() when omitted)
  streamFn: models.streamSimple.bind(models),

  // Session ID for provider caching
  sessionId: "session-123",

  // Dynamic API key resolution (for expiring OAuth tokens)
  getApiKey: async (provider) => refreshToken(),

  // Tool execution mode: "parallel" (default) or "sequential"
  toolExecution: "parallel",

  // Preflight each tool call after args are validated. Can block execution.
  beforeToolCall: async ({ toolCall, args, context }) => {
    if (toolCall.name === "bash") {
      return { block: true, reason: "bash is disabled", terminate: true };
    }
  },

  // Postprocess each tool result before final tool events are emitted.
  afterToolCall: async ({ toolCall, result, isError, context }) => {
    if (toolCall.name === "notify_done" && !isError) {
      return { terminate: true };
    }
    if (!isError) {
      return { details: { ...result.details, audited: true } };
    }
  },

  // Stop gracefully after a completed turn, before queued messages are polled.
  shouldStopAfterTurn: async ({ context }, signal) => {
    return shouldCompactBeforeNextTurn(context.messages);
  },

  // Replace context/model/thinking state before the next turn starts.
  prepareNextTurnWithContext: async ({ context, message }) => undefined,

  // Custom thinking budgets for token-based providers
  thinkingBudgets: {
    minimal: 128,
    low: 512,
    medium: 1024,
    high: 2048,
  },
});

The Agent class accepts shouldStopAfterTurn in AgentOptions. Low-level loop callers set the same hook in AgentLoopConfig. It runs after turn_end is emitted and after the assistant response and any tool executions have completed normally. If it returns true, the loop emits agent_end and exits before polling steering or follow-up queues, and before starting another LLM call. It does not abort the provider stream, does not cancel running tools, and does not alter the assistant message stop reason. The AgentOptions callback also receives the active run's AbortSignal as its second argument.

Agent State

interface AgentState {
  systemPrompt: string;
  model: Model<any>;
  thinkingLevel: ThinkingLevel;
  tools: AgentTool<any>[];
  messages: AgentMessage[];
  readonly isStreaming: boolean;
  readonly streamingMessage?: AgentMessage;
  readonly pendingToolCalls: ReadonlySet<string>;
  readonly errorMessage?: string;
}

Access state via agent.state.

Assigning agent.state.tools = [...] or agent.state.messages = [...] copies the top-level array before storing it. Mutating the returned array mutates the current agent state.

During streaming, agent.state.streamingMessage contains the current partial assistant message.

agent.state.isStreaming remains true until the run fully settles, including awaited agent_end subscribers. agent.state.errorMessage holds the error message from the most recent failed or aborted assistant turn, if any.

Methods

Prompting

// Text prompt
await agent.prompt("Hello");

// With images
await agent.prompt("What's in this image?", [
  { type: "image", data: base64Data, mimeType: "image/jpeg" }
]);

// AgentMessage directly
await agent.prompt({ role: "user", content: [{ type: "text", text: "Hello" }], timestamp: Date.now() });

// Batch of messages
await agent.prompt([msg1, msg2]);

// Continue from current context (last message must be user or toolResult)
await agent.continue();

Only one run at a time: prompt() / continue() throw while a run is active. Use steer() / followUp() to queue messages, or wait for completion. reset() throws while a run is active.

State Management

agent.state.systemPrompt = "New prompt";
agent.state.model = getModel("openai", "gpt-4o");
agent.state.thinkingLevel = "medium";
agent.state.tools = [myTool];
agent.toolExecution = "sequential";
agent.beforeToolCall = async ({ toolCall }) => undefined;
agent.afterToolCall = async ({ toolCall, result }) => undefined;
agent.shouldStopAfterTurn = async ({ context }) => shouldCompactBeforeNextTurn(context.messages);
agent.state.messages = newMessages; // top-level array is copied
agent.state.messages.push(message);
agent.reset(); // clears transcript, runtime state, and both queues

Session and Thinking Budgets

agent.sessionId = "session-123";

agent.thinkingBudgets = {
  minimal: 128,
  low: 512,
  medium: 1024,
  high: 2048,
};

Control

agent.abort();             // Cancel current operation
await agent.waitForIdle(); // Wait for completion (settles after awaited agent_end listeners)
agent.signal;              // Active run AbortSignal, if any
agent.hasQueuedMessages(); // True when either queue still holds messages

Events

const unsubscribe = agent.subscribe(async (event, signal) => {
  if (event.type === "agent_end") {
    // Final barrier work for the run
    await flushSessionState(signal);
  }
});
unsubscribe();

Steering and Follow-up

Steering messages let you interrupt the agent while tools are running. Follow-up messages let you queue work after the agent would otherwise stop.

agent.steeringMode = "one-at-a-time";
agent.followUpMode = "one-at-a-time";

// While agent is running tools
agent.steer({
  role: "user",
  content: [{ type: "text", text: "Stop! Do this instead." }],
  timestamp: Date.now(),
});

// After the agent finishes its current work
agent.followUp({
  role: "user",
  content: [{ type: "text", text: "Also summarize the result." }],
  timestamp: Date.now(),
});

const steeringMode = agent.steeringMode;
const followUpMode = agent.followUpMode;

agent.clearSteeringQueue();
agent.clearFollowUpQueue();
agent.clearAllQueues();

Use clearSteeringQueue, clearFollowUpQueue, or clearAllQueues to drop queued messages.

When steering messages are detected after a turn completes:

1. All tool calls from the current assistant message have already finished
2. Steering messages are injected
3. The LLM responds on the next turn

Follow-up messages are checked only when there are no more tool calls and no steering messages. If any are queued, they are injected and another turn runs.

Queue modes: "one-at-a-time" (default) drains only the oldest queued message per drain point; "all" drains every queued message at that point.

Custom Message Types

Extend AgentMessage via declaration merging:

declare module "@m4ike1/ion-agent-core" {
  interface CustomAgentMessages {
    notification: { role: "notification"; text: string; timestamp: number };
  }
}

// Now valid
const msg: AgentMessage = { role: "notification", text: "Info", timestamp: Date.now() };

Handle custom types in convertToLlm:

const agent = new Agent({
  streamFn: models.streamSimple.bind(models),
  convertToLlm: (messages) => messages.flatMap(m => {
    if (m.role === "notification") return []; // Filter out
    return [m];
  }),
});

Tools

Define tools using AgentTool:

import { Type } from "typebox";

const readFileTool: AgentTool = {
  name: "read_file",
  label: "Read File",  // For UI display
  description: "Read a file's contents",
  parameters: Type.Object({
    path: Type.String({ description: "File path" }),
  }),
  // Override execution mode for this tool (optional).
  // "sequential" forces the entire batch to run one at a time.
  // "parallel" allows concurrent execution with other tool calls.
  // If omitted, the global toolExecution config applies.
  executionMode: "sequential",
  execute: async (toolCallId, params, signal, onUpdate) => {
    const content = await fs.readFile(params.path, "utf-8");

    // Optional: stream progress
    onUpdate?.({ content: [{ type: "text", text: "Reading..." }], details: {} });

    // Optional: add `terminate: true` here to skip the automatic follow-up LLM call
    // when every finalized tool result in the batch does the same.
    return {
      content: [{ type: "text", text: content }],
      details: { path: params.path, size: content.length },
    };
  },
};

agent.state.tools = [readFileTool];

AgentTool also supports prepareArguments (compatibility shim normalizing raw tool-call arguments to the schema before validation) and replay ("never" | "safe", recovery policy for a durable effect whose outcome is unknown).

Error Handling

Throw an error when a tool fails. Do not return error messages as content.

execute: async (toolCallId, params, signal, onUpdate) => {
  if (!fs.existsSync(params.path)) {
    throw new Error(`File not found: ${params.path}`);
  }
  // Return content only on success
  return { content: [{ type: "text", text: "..." }] };
}

Thrown errors are caught by the agent and reported to the LLM as tool errors with isError: true.

Return terminate: true from execute(), a blocked beforeToolCall, or afterToolCall to hint that the agent should stop after the current tool batch. This only takes effect when every finalized tool result in the batch is terminating. The hint is runtime-only; emitted toolResult transcript messages remain standard LLM tool results.

Proxy Usage

For browser apps that proxy through a backend:

import { Agent, streamProxy } from "@m4ike1/ion-agent-core";

const agent = new Agent({
  streamFn: (model, context, options) =>
    streamProxy(model, context, {
      ...options,
      authToken: "...",
      proxyUrl: "https://your-server.com",
    }),
});

streamProxy POSTs { model, context, options } to ${proxyUrl}/api/stream with a Bearer auth token and reconstructs the assistant stream client-side. Only serializable stream options are forwarded (temperature, samplingParams, maxTokens, reasoning, cacheRetention, sessionId, headers, metadata, transport, thinkingBudgets, maxRetryDelayMs).

Low-Level API

For direct control without the Agent class:

import { agentLoop, agentLoopContinue } from "@m4ike1/ion-agent-core";

const context: AgentContext = {
  systemPrompt: "You are helpful.",
  messages: [],
  tools: [],
};

const config: AgentLoopConfig = {
  model: getModel("openai", "gpt-4o"),
  convertToLlm: (msgs) => msgs.filter(m => ["user", "assistant", "toolResult"].includes(m.role)),
  toolExecution: "parallel",  // overridden by per-tool executionMode if set
  beforeToolCall: async ({ toolCall, args, context }) => undefined,
  afterToolCall: async ({ toolCall, result, isError, context }) => undefined,
  shouldStopAfterTurn: async ({ message, toolResults, context, newMessages }) => {
    return shouldCompactBeforeNextTurn(context.messages);
  },
};

const userMessage = { role: "user", content: [{ type: "text", text: "Hello" }], timestamp: Date.now() };

const streamFn = models.streamSimple.bind(models);
for await (const event of agentLoop([userMessage], context, config, undefined, streamFn)) {
  console.log(event.type);
}

// Continue from existing context
for await (const event of agentLoopContinue(context, config, undefined, streamFn)) {
  console.log(event.type);
}

runAgentLoop / runAgentLoopContinue are the promise-based variants: same arguments plus an emit sink, resolving to the new messages.

These low-level streams are observational. They preserve event order, but they do not wait for your async event handling to settle before later producer phases continue. If you need message processing to act as a barrier before tool preflight, use the Agent class instead of raw agentLoop() or agentLoopContinue().

License

MIT