pi-mono/packages/coding-agent/test/suite/harness.ts
Armin Ronacher 8562bcf66a feat(coding-agent): codemode and MCP
This adds codemode and MCP support to pi. Codemode (packages/codemode) runs
model-written JavaScript in a QuickJS wasm VM inside a worker. From there the
script calls pi's tools as async functions, uses a per-session store, and can
read the model catalog and run classifiers. The codemode tool in coding-agent
is compatible with Codex's exec tool, and codemode.mode works like Codex's
tool modes (on or only). Large results are written to a JSON temp file. MCP
(packages/mcp) is a standalone client for stdio and streamable HTTP servers,
with OAuth.

The core only gets general mechanisms; codemode, tool_search and MCP are
built-in extensions that use them. Other extensions can replace these
built-ins. Tools declare an exposure: direct, model-only, codemode, deferred
or hidden. prepareLoadout() lets an orchestrating tool change which tool
declarations the model sees. ctx.executeTool() runs nested calls through
validation and the tool_call/tool_result hooks. Those calls emit events with
parentToolCallId and are recorded as bounded nestedCalls on the parent
result, which compaction and HTML export use. Tools can also return
structuredContent with an outputSchema, and can report errors with isError
instead of throwing. MCP servers come from mcp.json (global, or per project
after the project is trusted) or from pi.registerMcpServer(). They are
managed with /mcp and pi mcp list|login|logout, and their runtimes load on
first use.

Closes #10040
2026-09-29 09:12:50 +02:00

273 lines
9 KiB
TypeScript

import { createInMemoryModelRegistry, createModelRegistry, getModelRuntime } from "../model-runtime-test-utils.ts";
/**
* Local test harness for the new coding-agent test suite.
*/
import { existsSync, mkdirSync, rmSync, writeFileSync } from "node:fs";
import { tmpdir } from "node:os";
import { join } from "node:path";
import type { AgentMessage, AgentTool } from "@earendil-works/pi-agent-core";
import { Agent } from "@earendil-works/pi-agent-core";
import type {
FauxModelDefinition,
FauxProviderRegistration,
FauxResponseStep,
Model,
ToolResultMessage,
} from "@earendil-works/pi-ai/compat";
import { registerFauxProvider, streamSimple } from "@earendil-works/pi-ai/compat";
import { AgentSession, type AgentSessionEvent } from "../../src/core/agent-session.ts";
import { AuthStorage } from "../../src/core/auth-storage.ts";
import type { ExtensionRunner, ExtensionUIContext } from "../../src/core/extensions/index.ts";
import { convertToLlm } from "../../src/core/messages.ts";
import { SessionManager } from "../../src/core/session-manager.ts";
import type { Settings } from "../../src/core/settings-manager.ts";
import { SettingsManager } from "../../src/core/settings-manager.ts";
import type { InlineExtension, ResourceLoader } from "../../src/index.ts";
import { theme } from "../../src/modes/interactive/theme/theme.ts";
import {
type CreateTestExtensionsResultInput,
createTestExtensionsResult,
createTestResourceLoader,
} from "../utilities.ts";
type MessageTextPart = { type: "text"; text: string };
export function getMessageText(message: unknown): string {
if (!message || typeof message !== "object" || !("content" in message)) {
return "";
}
const content = (message as { content?: string | Array<{ type: string; text?: string }> }).content;
if (content === undefined) {
return "";
}
if (typeof content === "string") {
return content;
}
return content
.filter((part): part is MessageTextPart => part.type === "text")
.map((part) => part.text)
.join("\n");
}
export function getUserTexts(harness: Harness): string[] {
return harness.session.messages
.filter((message) => message.role === "user")
.map((message) => getMessageText(message));
}
export function getAssistantTexts(harness: Harness): string[] {
return harness.session.messages
.filter((message) => message.role === "assistant")
.map((message) => getMessageText(message));
}
/** The latest result of `toolName` in the session transcript. */
export function getToolResult(harness: Harness, toolName: string): ToolResultMessage {
const result = harness.session.messages.findLast(
(message): message is ToolResultMessage => message.role === "toolResult" && message.toolName === toolName,
);
if (!result) throw new Error(`No ${toolName} tool result`);
return result;
}
/** An extension UI context that does nothing, with `overrides` applied. */
export function createTestUiContext(overrides: Partial<ExtensionUIContext> = {}): ExtensionUIContext {
return {
select: async () => undefined,
confirm: async () => false,
input: async () => undefined,
notify: () => {},
onTerminalInput: () => () => {},
setStatus: () => {},
setWorkingMessage: () => {},
setWorkingVisible: () => {},
setWorkingIndicator: () => {},
setHiddenThinkingLabel: () => {},
setWidget: () => {},
setFooter: () => {},
setHeader: () => {},
setTitle: () => {},
custom: async <T>() => undefined as T,
pasteToEditor: () => {},
setEditorText: () => {},
getEditorText: () => "",
editor: async () => undefined,
addAutocompleteProvider: () => {},
setEditorComponent: () => {},
getEditorComponent: () => undefined,
get theme() {
return theme;
},
getAllThemes: () => [],
getTheme: () => undefined,
setTheme: () => ({ success: false, error: "Theme switching not available in tests" }),
getToolsExpanded: () => false,
setToolsExpanded: () => {},
...overrides,
};
}
export interface HarnessOptions {
models?: FauxModelDefinition[];
settings?: Partial<Settings>;
tools?: AgentTool[];
initialActiveToolNames?: string[];
allowedToolNames?: string[];
excludedToolNames?: string[];
resourceLoader?: ResourceLoader;
extensionFactories?: Array<InlineExtension | CreateTestExtensionsResultInput>;
withConfiguredAuth?: boolean;
modelsJson?: Record<string, unknown>;
}
export interface Harness {
session: AgentSession;
sessionManager: SessionManager;
settingsManager: SettingsManager;
authStorage: AuthStorage;
faux: FauxProviderRegistration;
models: [Model<string>, ...Model<string>[]];
getModel(): Model<string>;
getModel(modelId: string): Model<string> | undefined;
setResponses: (responses: FauxResponseStep[]) => void;
appendResponses: (responses: FauxResponseStep[]) => void;
getPendingResponseCount: () => number;
events: AgentSessionEvent[];
eventsOfType<T extends AgentSessionEvent["type"]>(type: T): Extract<AgentSessionEvent, { type: T }>[];
tempDir: string;
cleanup: () => void;
}
function createTempDir(): string {
const tempDir = join(tmpdir(), `pi-suite-${Date.now()}-${Math.random().toString(36).slice(2)}`);
mkdirSync(tempDir, { recursive: true });
return tempDir;
}
export async function createHarness(options: HarnessOptions = {}): Promise<Harness> {
const tempDir = createTempDir();
const fauxProvider: FauxProviderRegistration = registerFauxProvider({
models: options.models,
});
fauxProvider.setResponses([]);
const model = fauxProvider.getModel();
const toolMap = options.tools ? Object.fromEntries(options.tools.map((tool) => [tool.name, tool])) : undefined;
const withConfiguredAuth = options.withConfiguredAuth ?? true;
const extensionRunnerRef: { current?: ExtensionRunner } = {};
const sessionManager = SessionManager.inMemory();
const settingsManager = SettingsManager.inMemory(options.settings);
const authStorage = AuthStorage.inMemory();
if (withConfiguredAuth) {
await authStorage.modify(model.provider, async () => ({ type: "api_key", key: "faux-key" }));
}
const modelsPath = options.modelsJson === undefined ? undefined : join(tempDir, "models.json");
if (modelsPath) writeFileSync(modelsPath, JSON.stringify(options.modelsJson));
const modelRegistry = modelsPath
? await createModelRegistry(authStorage, modelsPath)
: await createInMemoryModelRegistry(authStorage);
if (withConfiguredAuth) {
modelRegistry.registerProvider(model.provider, {
baseUrl: model.baseUrl,
apiKey: "faux-key",
api: fauxProvider.api,
models: fauxProvider.models.map((registeredModel) => ({
id: registeredModel.id,
name: registeredModel.name,
api: registeredModel.api,
reasoning: registeredModel.reasoning,
input: registeredModel.input,
inputLimits: registeredModel.inputLimits,
cost: registeredModel.cost,
contextWindow: registeredModel.contextWindow,
maxTokens: registeredModel.maxTokens,
baseUrl: registeredModel.baseUrl,
})),
});
}
const agent = new Agent({
getApiKey: () => (withConfiguredAuth ? "faux-key" : undefined),
streamFn: streamSimple,
initialState: {
model,
systemPrompt: "",
tools: [],
},
convertToLlm,
onPayload: async (payload) => {
const runner = extensionRunnerRef.current;
if (!runner?.hasHandlers("before_provider_request")) {
return payload;
}
return runner.emitBeforeProviderRequest(payload);
},
onResponse: async (response) => {
const runner = extensionRunnerRef.current;
if (!runner?.hasHandlers("after_provider_response")) {
return;
}
await runner.emit({
type: "after_provider_response",
status: response.status,
headers: response.headers,
});
},
transformContext: async (messages: AgentMessage[]) => {
const runner = extensionRunnerRef.current;
if (!runner) return messages;
return runner.emitContext(messages);
},
});
const extensionsResult = options.extensionFactories
? await createTestExtensionsResult(options.extensionFactories, tempDir)
: undefined;
const resourceLoader =
options.resourceLoader ?? createTestResourceLoader(extensionsResult ? { extensionsResult } : undefined);
const session = new AgentSession({
agent,
sessionManager,
settingsManager,
cwd: tempDir,
modelRuntime: getModelRuntime(modelRegistry),
resourceLoader,
baseToolsOverride: toolMap,
initialActiveToolNames: options.initialActiveToolNames,
allowedToolNames: options.allowedToolNames,
excludedToolNames: options.excludedToolNames,
extensionRunnerRef,
});
const events: AgentSessionEvent[] = [];
session.subscribe((event) => {
events.push(event);
});
return {
session,
sessionManager,
settingsManager,
authStorage,
faux: fauxProvider,
models: fauxProvider.models,
getModel: fauxProvider.getModel,
setResponses: fauxProvider.setResponses,
appendResponses: fauxProvider.appendResponses,
getPendingResponseCount: fauxProvider.getPendingResponseCount,
events,
eventsOfType<T extends AgentSessionEvent["type"]>(type: T) {
return events.filter((event): event is Extract<AgentSessionEvent, { type: T }> => event.type === type);
},
tempDir,
cleanup() {
session.dispose();
fauxProvider.unregister();
if (existsSync(tempDir)) {
rmSync(tempDir, { recursive: true });
}
},
};
}