Skip to main content
An agent run represents a single execution of your agent, from the moment it receives a user input to the final response. All setter methods return this, so you can chain calls: 
r.prompt("What is the weather in London?").metadata({ userId: "user-123" }).response("15°C and cloudy.");
await r.end();

​
Required

​
Initialize run

import { run } from "@contextcompany/custom";

const r = run();
ParameterRequiredDescription
runIdNoCustom run ID. If not provided, a UUID is automatically generated.
sessionIdNoGroup multiple runs into an agent session. See Track agent sessions.
conversationalNoSet to true to indicate the run was initiated by a user. See Conversational runs.
startTimeNoCustom start time. Defaults to new Date().
timeoutNoTimeout in milliseconds. Defaults to 1200000 (20 minutes).
import { run } from "@contextcompany/custom";

const r = run({
  runId: "your-custom-uuid",
  sessionId: "session-123",
  conversational: true,
});
Every run has a .runId property you can use to retrieve its ID (auto-generated or custom): 
const runId = r.runId;

​
Set prompt

The prompt is required before calling .end(). 
r.prompt(userPrompt);
// or
r.prompt({ user_prompt: userPrompt, system_prompt: systemPrompt });
ParameterRequiredDescription
userPromptYesThe user input to the agent
system_promptNoA system message to include alongside the user prompt
// User prompt only
r.prompt("What is the weather in London?");

// User prompt with system prompt
r.prompt({
  user_prompt: "What is the weather in London?",
  system_prompt: "You are a helpful weather agent.",
});

​
End run

End the run and export it. A run must have a prompt set before calling .end(). The run cannot be modified after calling .end(). 
import { run } from "@contextcompany/custom";

const r = run();
r.prompt("What is the weather in London?");

// --- Your agent loop ---
// ...
// --- End of agent loop ---

r.response("The weather in London is 15°C and cloudy.");
await r.end();

​
Optional run data

​
Set response

r.response(text);
ParameterRequiredDescription
textYesThe final response from the agent
r.response("The weather in London is 15°C and cloudy.");

​
Mark run as failed

Sets the status code to error and exports the payload. The run cannot be modified after calling .error(). Unlike .end(), calling .error() does not require a prompt to be set. Any un-ended children (steps, tool calls) are automatically ended with error status. 
r.error(statusMessage?);
ParameterRequiredDescription
statusMessageNoA description of the error
import { run } from "@contextcompany/custom";

const r = run();
r.prompt("What is the weather in London?");

try {
  const result = await callWeatherApi();
  r.response(result);
  await r.end();
} catch (e) {
  await r.error(String(e));
}

​
Add custom metadata

Custom metadata allows you to add additional properties to your agent runs. This is particularly useful for tying agent runs to your own specific business logic, letting you filter and analyze agent runs by user, organization, feature, or some other dimension. 
r.metadata(data);
ParameterRequiredDescription
dataYesAn object of key-value pairs
r.metadata({
  userId: "4a6b111c-b53a-4d00-a877-67185022ab9e",
  orgId: "org-123",
});
Agent runs are automatically indexed by your custom metadata fields and can be filtered directly in the dashboard.

​
Identifying the agent

If your product ships more than one named agent, set the reserved tcc.agent metadata key to scope the run to a specific agent. The dashboard’s top-level agent selector, per-agent patterns and recaps, and the agent filter on the REST API and MCP tools all read from this key. 
r.metadata({
  "tcc.agent": "support-agent",
});
Agent names that collide with reserved dashboard routes (for example runs, sessions, patterns, recaps, overview, search, failures, feedback, tools, topics, views, settings, mcp-and-api) are dropped.

​
Identifying users and organizations

Attach the end user and their organization to a run as first-class identity using the reserved tcc.userId, tcc.userName, tcc.orgId, and tcc.orgName metadata keys. This is not the same as adding a userId field to custom metadata — these keys promote user and org identity to dedicated dashboard filters and unlock native user/org search, per-user views, and per-org analytics. See User and organization identity for the full concept. Set these whenever you have a stable identifier for the end user or their organization in your product. 
r.metadata({
  "tcc.userId": "user-123",
  "tcc.userName": "Jane Doe",
  "tcc.orgId": "org-456",
  "tcc.orgName": "Acme Inc.",
});
tcc.userName and tcc.orgName require the corresponding ID (tcc.userId / tcc.orgId) to also be set. Names without IDs are dropped.

​
Add user feedback

User feedback allows you to collect score (thumbs up & thumbs down) and text feedback (up to 2000 characters) from end users on your agent runs. This is useful for tracking user satisfaction, identifying problematic responses, and filtering agent runs in the dashboard to focus on positive or negative feedback. Both score and text are optional individually, but each request must include at least one of them. score is the thumbs rating. Use only "thumbs_up" or "thumbs_down". text is written feedback from your user, up to 2000 characters.
ParameterRequiredDescription
scoreNoThumbs rating. Must be "thumbs_up" or "thumbs_down"
textNoWritten user feedback, up to 2000 characters
Use the submitFeedback function with a run ID: 
import { run, submitFeedback } from "@contextcompany/custom";

const r = run();
r.prompt("What is the weather in London?");
r.response("The weather in London is 15°C and cloudy.");
await r.end();

// Submit score and/or text feedback using the run ID
await submitFeedback({
  runId: r.runId,
  score: "thumbs_up", // Optional thumbs rating: "thumbs_up" or "thumbs_down"
  text: "Very helpful!", // Optional written user feedback, up to 2000 characters
});
Agent runs with feedback can be filtered in the dashboard using the feedback filter.

​
Track agent sessions

Agent sessions represent multiple agent runs that are grouped together. The most common use case is tracking entire conversations between a human user and an AI agent in chatbot interfaces. Agent sessions can be tracked by passing a sessionId when creating a run: 
import { run } from "@contextcompany/custom";

const sessionId = "unique-session-identifier";

// First message in the conversation
const r1 = run({ sessionId });
r1.prompt("What is the weather in London?");
r1.response("The weather in London is 15°C and cloudy.");
await r1.end();

// Second message in the same conversation
const r2 = run({ sessionId });
r2.prompt("What about Paris?");
r2.response("The weather in Paris is 18°C and sunny.");
await r2.end();
The sessionId should be a unique string identifier. We recommend using a UUID. Agent sessions are automatically indexed and can be filtered directly in the dashboard.

​
Factory pattern

If you have all run data available upfront (post-hoc logging, batch imports, replaying from logs), use the sendRun function: 
import { sendRun } from "@contextcompany/custom";

await sendRun({
  prompt: {
    user_prompt: "What's the weather?",
    system_prompt: "You are a helpful assistant.",
  },
  response: "72°F in San Francisco",
  startTime: new Date("2025-01-01T00:00:00Z"),
  endTime: new Date("2025-01-01T00:00:01Z"),
  metadata: { agent: "weather-bot" },
  steps: [
    {
      prompt: JSON.stringify(messages),
      response: assistantContent,
      model: "gpt-4o",
      tokens: { uncached: 120, completion: 45 },
      startTime: new Date("2025-01-01T00:00:00Z"),
      endTime: new Date("2025-01-01T00:00:01Z"),
    },
  ],
  toolCalls: [
    {
      name: "get_weather",
      args: { city: "San Francisco" },
      result: { temp: 72, unit: "F" },
      startTime: new Date("2025-01-01T00:00:00Z"),
      endTime: new Date("2025-01-01T00:00:01Z"),
    },
  ],
});