A live forecast of the context window, drawn above the prompt.

English | 中文
Karaka is a self-hosted server for adding persistent AI agents to an application. Connect a backend or browser frontend to named agents, stream their responses, and let them call selected application functions as tools.
Each named agent is an Agent Preset: a Cordis plugin composition that defines its prompt, tools, skills, and model behavior. Karaka runs as a separate process and keeps durable chat state; your application owns user authentication and business authorization.
Application backend Karaka process
@karaka-ai/sdk chat client -- HTTP / SSE --> named Agent Preset
@karaka-ai/sdk MCP tools <-- MCP -- selected application tools
Browser frontend Karaka process
@karaka-ai/agent/browser -- HTTP / SSE --> authenticated chat and events
Backend integrations use @karaka-ai/sdk for chat over HTTP/SSE and for hosting authenticated MCP tools on the application's existing HTTP server. The backend supplies trusted tenant and user identifiers. Karaka binds those identifiers to the chat and forwards them to tool callbacks, where the application enforces business authorization. Installing the SDK starts no process and opens no port.
Browser integrations use @karaka-ai/agent/browser with expiring credentials issued by the application backend. Each credential binds the application, tenant, and user identity; browser requests cannot choose another identity. See the browser client setup for authentication and server configuration.
<a id="run"></a>
Read the runtime implementation guide for composition, builds, and deployment limitations. The CLI and SDK are maintained in their own repositories.
Karaka builds on the open-source DeepSeek Harness developed by DeepSeek AI and retains its everything-is-a-plugin architecture powered by Cordis. Karaka has its own @karaka-ai/* packages and karaka CLI.
Karaka and its inherited Harness runtime are in developer preview and may make compatibility-breaking changes. Review the safety notice before running the project.
<a id="run-from-source"></a>
Start with the development guide and architecture documentation.
See CONTRIBUTING.md before proposing a change.
For agents, follow AGENTS.md.
Third-party dependencies and their licenses are disclosed in THIRD_PARTY_NOTICES.md.
hooks/token-weather.mjs 88 lines1// Token Weather, from "Getting started with Claude Code mods"
2// (https://claude.dev/blog/getting-started-with-claude-code-mods/, Anthropic, 2026-10-01).
3// The published module, unchanged.
4
5// Token Weather: a live forecast of the context window, above the prompt.
6
7const HISTORY = 12;
8const BARS = "▁▂▃▄▅▆▇█";
9const FORECAST = [
10 { upTo: 25, icon: "☀", word: "Clear", color: "yellow" },
11 { upTo: 50, icon: "☁", word: "Cloudy", color: "cyan" },
12 { upTo: 75, icon: "☂", word: "Showers", color: "blue" },
13 { upTo: 90, icon: "☇", word: "Storm", color: "magenta" },
14 { upTo: Infinity, icon: "↯", word: "Compact soon", color: "red" },
15];
16
17// Held by the host, so the history survives a hot reload of this file.
18const readings = { plugin: "token-weather", key: "readings" };
19
20export function register(on) {
21 on("session.start", async ($, e, next) => {
22 const result = await next(e);
23 await takeReading($);
24 return result;
25 });
26
27 on("turn.complete", async ($, e, next) => {
28 const result = await next(e);
29 if (!e.agentId) {
30 await takeReading($); // main-loop turns only, not subagents
31 }
32 return result;
33 });
34
35 on("ui.render", { component: "AbovePrompt" }, async ($, e, next) => {
36 const { value: history = [] } = await $.state.get(readings);
37 if (e.props.hasSurvey || history.length === 0) {
38 return next(e);
39 }
40 const { Box, Text } = $.ui.resolve(e);
41 return band(Box, Text, history, e.props.bodyColumns);
42 });
43}
44
45async function takeReading($) {
46 const { context } = await $.session.usage();
47 if (!context?.window) return;
48 const tokens = context.tokens ?? 0;
49 const percent = context.percent ?? Math.round((tokens / context.window) * 100);
50 const { value: history = [] } = await $.state.get(readings);
51 await $.state.set(readings, [...history, { tokens, window: context.window, percent }].slice(-HISTORY));
52}
53
54function band(Box, Text, history, columns) {
55 const now = history[history.length - 1];
56 const f = FORECAST.find((b) => now.percent < b.upTo);
57 const parts = [
58 Text({ color: f.color, bold: true, children: `${f.icon} ${f.word}` }),
59 Text({ children: ` ${now.percent}% of context` }),
60 Text({ dimColor: true, children: ` ${short(now.tokens)} / ${short(now.window)}` }),
61 ];
62 if (columns >= 60) {
63 parts.push(Text({ dimColor: true, children: " last turns " }));
64 parts.push(Text({ color: f.color, children: sparkline(history) }));
65 if (history.length > 1) {
66 parts.push(Text({ dimColor: true, children: trend(history) }));
67 }
68 }
69 return Box({ flexDirection: "row", paddingX: 1, children: parts });
70}
71
72function sparkline(history) {
73 const top = Math.max(...history.map((r) => r.tokens), 1);
74 return history.map((r) => BARS[Math.floor((r.tokens / top) * (BARS.length - 1))]).join("");
75}
76
77function trend(history) {
78 const delta = history[history.length - 1].tokens - history[history.length - 2].tokens;
79 if (delta === 0) return " steady";
80 return delta > 0 ? ` ▲ +${short(delta)} last turn` : ` ▼ ${short(-delta)} last turn`;
81}
82
83function short(n) {
84 if (n >= 1_000_000) return `${+(n / 1_000_000).toFixed(1)}M`;
85 if (n >= 1_000) return `${+(n / 1_000).toFixed(1)}k`;
86 return String(n);
87}
88types/index.d.ts 11 lines1// Token Weather's type contract, from https://claude.dev/blog/getting-started-with-claude-code-mods/
2// (Anthropic, 2026-10-01). The published file, unchanged.
3
4export type TokenWeatherReading = { tokens: number; window: number; percent: number }
5
6declare module 'claude-code' {
7 interface PluginState {
8 'token-weather': { readings: TokenWeatherReading[] }
9 }
10}
11