Skip to content

nanocoai/nanoclaw / .claude/skills/add-codex

add-codex

Flagged

Use Codex (OpenAI's codex app-server) as a full agent provider — planning, tool orchestration, MCP tools, server-side history, session resume — alongside or instead of Claude. ChatGPT subscription or OpenAI API key, vault-only via the selected gateway. Per-group via `ncl groups config update --provider codex`. Distinct from using OpenAI as an MCP tool (where Claude remains the planner).

payload/container/agent-runner/src/providers/codex.factory.test.ts

View on GitHub
import { describe, expect, it } from 'bun:test';

import { type AppServer, codexRuntimeOwnership, type writeCodexConfigToml } from './codex-app-server.js';
import { CodexProvider, type CodexRuntimeDeps } from './codex.js';

const MEMORY_HOOK = { command: 'bun /app/src/memory/hook.ts', legacyCommands: [], sources: ['startup'] };

/**
 * Runtime deps that record config writes and stop the query at the app-server
 * handshake, so no codex binary is ever spawned.
 */
function recordingRuntime(): { runtime: CodexRuntimeDeps; writes: Parameters<typeof writeCodexConfigToml>[] } {
  const writes: Parameters<typeof writeCodexConfigToml>[] = [];
  const runtime: CodexRuntimeDeps = {
    writeCodexConfigToml: (...args) => {
      writes.push(args);
    },
    spawnCodexAppServer: () =>
      ({
        process: { stdin: { write: () => true } },
        readline: { close: () => {} },
        pending: new Map(),
        notificationHandlers: [],
        exitHandlers: [],
        serverRequestHandlers: [],
      }) as unknown as AppServer,
    attachCodexAutoApproval: () => {},
    initializeCodexAppServer: async () => {
      throw new Error('handshake stopped by test');
    },
    startOrResumeCodexThread: async () => 'thread-unreached',
    startCodexTurn: async () => 'turn-unreached',
    steerCodexTurn: async () => {},
    interruptCodexTurn: async () => {},
    killCodexAppServer: () => {},
  };
  return { runtime, writes };
}

async function runQueryToHandshake(provider: CodexProvider): Promise<void> {
  provider.registerMemorySessionHook(MEMORY_HOOK);
  const query = provider.query({ prompt: 'hello', cwd: '/workspace/agent' });
  await expect(query.events.next()).rejects.toThrow('handshake stopped by test');
}

describe('CodexProvider', () => {
  it('rejects unsupported reasoning effort values', () => {
    expect(() => new CodexProvider({ effort: 'max' })).toThrow(/Unsupported Codex reasoning effort/);
  });

  it('normalizes supported reasoning effort values', () => {
    expect(new CodexProvider({ effort: 'HIGH' })).toBeInstanceOf(CodexProvider);
  });

  it('accepts supported reasoning effort values', () => {
    expect(new CodexProvider({ effort: 'xhigh' })).toBeInstanceOf(CodexProvider);
  });

  it('requires the shared memory hook before starting a query', () => {
    expect(() => new CodexProvider({}).query({ prompt: 'hello', cwd: '/workspace/agent' })).toThrow(/not registered/);
  });

  it('writes the runtime files exactly once per query when no contract owns them', async () => {
    const previous = codexRuntimeOwnership.contractOwnsRuntimeFiles;
    codexRuntimeOwnership.contractOwnsRuntimeFiles = false;
    try {
      const { runtime, writes } = recordingRuntime();
      const servers = { nanoclaw: { command: 'bun' } };
      await runQueryToHandshake(
        new CodexProvider({ mcpServers: servers, model: 'gpt-5', effort: 'HIGH', speed: 'fast' }, runtime),
      );
      expect(writes).toHaveLength(1);
      const [mcpServers, hook, inference] = writes[0];
      expect(mcpServers).toEqual(servers);
      expect(hook).toEqual(MEMORY_HOOK);
      // The same normalized values the contract path renders from.
      expect(inference).toEqual({ model: 'gpt-5', effort: 'high', fastMode: true });
    } finally {
      codexRuntimeOwnership.contractOwnsRuntimeFiles = previous;
    }
  });

  it('leaves the runtime files to the contract when its module owns them', async () => {
    const previous = codexRuntimeOwnership.contractOwnsRuntimeFiles;
    codexRuntimeOwnership.contractOwnsRuntimeFiles = true;
    try {
      const { runtime, writes } = recordingRuntime();
      await runQueryToHandshake(new CodexProvider({ model: 'gpt-5' }, runtime));
      expect(writes).toHaveLength(0);
    } finally {
      codexRuntimeOwnership.contractOwnsRuntimeFiles = previous;
    }
  });

  it('consumes core-resolved configuration instead of re-deriving it', async () => {
    const previous = codexRuntimeOwnership.contractOwnsRuntimeFiles;
    codexRuntimeOwnership.contractOwnsRuntimeFiles = false;
    try {
      const { runtime, writes } = recordingRuntime();
      const resolvedServers = { resolved: { command: 'resolved-bin' } };
      const provider = new CodexProvider({ mcpServers: { ignored: { command: 'x' } }, model: 'ignored' }, runtime, {
        executionPolicy: {},
        inference: { model: 'gpt-5', effort: 'high', fastMode: undefined },
        mcpServers: resolvedServers,
      });
      await runQueryToHandshake(provider);
      const [mcpServers, , inference] = writes[0];
      expect(mcpServers).toEqual(resolvedServers);
      expect(inference).toEqual({ model: 'gpt-5', effort: 'high', fastMode: undefined });
    } finally {
      codexRuntimeOwnership.contractOwnsRuntimeFiles = previous;
    }
  });

  it.each([undefined, 'pragmatic'])('passes resolved tone to thread settings (%j)', async (personality) => {
    const { runtime } = recordingRuntime();
    runtime.initializeCodexAppServer = async () => {};
    runtime.startOrResumeCodexThread = async (_server, _thread, settings) => {
      expect(settings.personality).toBe(personality ?? 'friendly');
      throw new Error('thread settings captured');
    };
    const provider = new CodexProvider({}, runtime, {
      inference: {},
      mcpServers: {},
      ...(personality ? { tone: { personality } } : {}),
    });
    provider.registerMemorySessionHook(MEMORY_HOOK);
    const query = provider.query({ prompt: 'hi', cwd: '/workspace/agent' });
    await expect(query.events.next()).rejects.toThrow('thread settings captured');
  });
});
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall