testronaut 1.6.1 → 1.7.0

package/tests/binTests/init.test.js

@@ -77,6 +77,9 @@ describe('initializeTestronautProject', () => {
     expect(cfg.initialized).toBe(true);
     expect(cfg.outputDir).toBe('missions/mission_reports');
     expect(typeof cfg.maxTurns).toBe('number');
+    expect(cfg.tags).toEqual([]);
+    expect(cfg.tagMatch).toBe('any');
+    expect(cfg.addTags).toEqual([]);
     expect(cfg.dom?.listItemLimit).toBe(3);
     expect(cfg.resourceGuard?.enabled).toBe(true);
     expect(cfg.humanInput?.enabled).toBe(true);
@@ -95,6 +98,9 @@ describe('initializeTestronautProject', () => {
     // Welcome mission (name depends on your implementation; we check presence in missions/)
     const files = fs.readdirSync(path.join(temp, 'missions'));
     expect(files.some(f => /welcome/i.test(f))).toBe(true);
+    const welcome = fs.readFileSync(path.join(temp, 'missions', 'welcome.mission.js'), 'utf8');
+    expect(welcome).toContain('export const tags = [];');
+    expect(welcome).toContain('tags\n  },');
 
     // Exactly two prompts were asked
     expect(shared.calls).toBe(2);
@@ -112,6 +118,9 @@ describe('initializeTestronautProject', () => {
     const cfg = readJson(path.join(temp, 'testronaut-config.json'));
     expect(cfg.provider).toBe('gemini');
     expect(cfg.model).toBe('gemini-2.5-flash');
+    expect(cfg.tags).toEqual([]);
+    expect(cfg.tagMatch).toBe('any');
+    expect(cfg.addTags).toEqual([]);
     expect(cfg.dom?.listItemLimit).toBe(3);
     expect(cfg.resourceGuard?.enabled).toBe(true);
     expect(cfg.humanInput?.enabled).toBe(true);
@@ -129,6 +138,21 @@ describe('initializeTestronautProject', () => {
     expect(shared.calls).toBe(2);
   });
 
+  it('scaffolds a new Anthropic project', async () => {
+    const temp = makeTempProject();
+    process.chdir(temp);
+    shared.answers.push({ llmProvider: 'anthropic' });
+    shared.answers.push({ anthropicModel: 'claude-sonnet-5' });
+
+    await initializeTestronautProject();
+
+    const cfg = readJson(path.join(temp, 'testronaut-config.json'));
+    expect(cfg.provider).toBe('anthropic');
+    expect(cfg.model).toBe('claude-sonnet-5');
+    expect(fs.readFileSync(path.join(temp, '.env'), 'utf8')).toMatch(/ANTHROPIC_API_KEY=sk-ant-/);
+    expect(shared.calls).toBe(2);
+  });
+
   it('is idempotent: when initialized, it skips prompts and does not overwrite .env', async () => {
     const temp = makeTempProject();
     process.chdir(temp);

package/tests/binTests/initHelpersTests/defaultConfig.test.js

@@ -8,6 +8,9 @@ describe('defaultConfig', () => {
     expect(cfg.outputDir).toBe('missions/mission_reports');
     expect(cfg.projectName).toBe('my-app');
     expect(cfg.maxTurns).toBe(20);
+    expect(cfg.tags).toEqual([]);
+    expect(cfg.tagMatch).toBe('any');
+    expect(cfg.addTags).toEqual([]);
     expect(cfg.dom?.listItemLimit).toBe(3);
     expect(cfg.resourceGuard?.enabled).toBe(true);
     expect(cfg.resourceGuard?.hrefIncludes).toContain('/document/');

package/tests/binTests/initHelpersTests/makeEnvTemplate.test.js

@@ -11,4 +11,8 @@ describe('makeEnvTemplate', () => {
     const s = makeEnvTemplate('gemini');
     expect(s).toMatch(/GEMINI_API_KEY=AIza\.\.\./);
   });
+
+  it('returns Anthropic template', () => {
+    expect(makeEnvTemplate('anthropic')).toMatch(/ANTHROPIC_API_KEY=sk-ant-/);
+  });
 });

package/tests/binTests/initHelpersTests/models.test.js

@@ -1,5 +1,5 @@
 import { describe, it, expect } from 'vitest';
-import { openAIModels, geminiModels, isKnownModel, pickInitialIndex } from '../../../bin/initHelpers';
+import { openAIModels, geminiModels, anthropicModels, isKnownModel, pickInitialIndex } from '../../../bin/initHelpers';
 
 describe('model helpers', () => {
   it('lists known models', () => {
@@ -16,11 +16,13 @@ describe('model helpers', () => {
     ]));
     expect(openAIModels()).not.toContain('gpt-5.3');
     expect(geminiModels()).toContain('gemini-2.5-flash');
+    expect(anthropicModels()).toContain('claude-sonnet-5');
   });
 
   it('recognizes known models by provider', () => {
     expect(isKnownModel('openai', 'gpt-4.1-mini')).toBe(true);
     expect(isKnownModel('gemini', 'gpt-4.1-mini')).toBe(false);
+    expect(isKnownModel('anthropic', 'claude-sonnet-5')).toBe(true);
   });
 
   it('picks initial index from list or fallback', () => {

package/tests/coreTests/missionDiscovery.test.js

@@ -31,11 +31,11 @@ describe('discoverMissionFiles', () => {
     fs.writeFileSync(path.join(missionsDir, 'beta.txt'), '// ignore');
     fs.mkdirSync(path.join(missionsDir, 'nested'), { recursive: true });
     fs.writeFileSync(path.join(missionsDir, 'nested', 'gamma.mission.js'), '// nested');
-    fs.writeFileSync(path.join(missionsDir, 'delta.mission.ts'), '// ts not included by default');
+    fs.writeFileSync(path.join(missionsDir, 'delta.mission.ts'), '// ts');
 
     const res = await discoverMissionFiles({ cwd: tmp });
     expect(res.root).toBe(path.resolve(tmp, 'missions'));
-    expect(res.files).toEqual(['alpha.mission.js']);
+    expect(res.files).toEqual(['alpha.mission.js', 'delta.mission.ts']);
   });
 
   it('respects missions.root with include patterns', async () => {

package/tests/coreTests/missionLoader.test.js

@@ -0,0 +1,52 @@
+import { afterEach, describe, expect, it } from 'vitest';
+import fs from 'fs';
+import os from 'os';
+import path from 'path';
+import { loadMissionModule } from '../../core/missionLoader.js';
+
+const tempDirs = [];
+
+function makeProject(packageJson) {
+  const dir = fs.mkdtempSync(path.join(os.tmpdir(), 'testronaut-loader-'));
+  tempDirs.push(dir);
+  if (packageJson) {
+    fs.writeFileSync(path.join(dir, 'package.json'), JSON.stringify(packageJson));
+  }
+  return dir;
+}
+
+afterEach(() => {
+  while (tempDirs.length) fs.rmSync(tempDirs.pop(), { recursive: true, force: true });
+});
+
+describe('loadMissionModule', () => {
+  it.each([
+    ['without a package type', undefined],
+    ['inside a CommonJS project', { type: 'commonjs' }],
+  ])('loads ESM-style .mission.js %s', async (_label, packageJson) => {
+    const dir = makeProject(packageJson);
+    const missionPath = path.join(dir, 'sample.mission.js');
+    fs.writeFileSync(missionPath, 'export const tags = ["smoke"]; export async function executeMission() { return "ok"; }');
+
+    const mission = await loadMissionModule(missionPath);
+
+    expect(mission.tags).toEqual(['smoke']);
+    await expect(mission.executeMission()).resolves.toBe('ok');
+  });
+
+  it('loads a typed .mission.ts file with relative imports', async () => {
+    const dir = makeProject({ type: 'commonjs' });
+    fs.writeFileSync(path.join(dir, 'goal.ts'), 'export const goal: string = "TypeScript works";');
+    const missionPath = path.join(dir, 'sample.mission.ts');
+    fs.writeFileSync(missionPath, [
+      'import { goal } from "./goal.ts";',
+      'export const tags: string[] = ["typescript"];',
+      'export async function executeMission(): Promise<string> { return goal; }',
+    ].join('\n'));
+
+    const mission = await loadMissionModule(missionPath);
+
+    expect(mission.tags).toEqual(['typescript']);
+    await expect(mission.executeMission()).resolves.toBe('TypeScript works');
+  });
+});

package/tests/coreTests/turnLoop.test.js

@@ -78,6 +78,7 @@ vi.mock('../../tools/toolSchema.js', () => ({
     { type: 'function', function: { name: 'screenshot', description: 'take a screenshot', parameters: {} } },
     { type: 'function', function: { name: 'list_local_files', description: 'list files', parameters: {} } },
     { type: 'function', function: { name: 'get_mfa_code', description: 'get mfa', parameters: {} } },
+    { type: 'function', function: { name: 'set_ground_control_state', description: 'set state', parameters: {} } },
   ],
 }));
 
@@ -126,6 +127,7 @@ describe('turnLoop', () => {
   let browser;
 
   beforeEach(() => {
+    delete process.env.TESTRONAUT_SCREENSHOTS;
     browser = {};
     // reset spies and chat mock
     shared.chromeToolSpies.fill.mockClear();
@@ -136,6 +138,19 @@ describe('turnLoop', () => {
     if (shared.chatMock) shared.chatMock.mockReset();
   });
 
+  it('does not expose the screenshot tool when screenshots are disabled', async () => {
+    process.env.TESTRONAUT_SCREENSHOTS = '0';
+    shared.chatMock.mockResolvedValueOnce({
+      message: { role: 'assistant', content: 'FINAL: done' },
+      usage: { total_tokens: 2 },
+    });
+
+    await turnLoop(browser, baseMessages(), 1, 0, 0, {}, { steps: [], missionName: 'demo' });
+
+    const tools = shared.chatMock.mock.calls[0][0].tools;
+    expect(tools.map(tool => tool.function.name)).not.toContain('screenshot');
+  });
+
   it('handles a final response in a single turn (no tools)', async () => {
     shared.chatMock.mockResolvedValueOnce({
       message: { role: 'assistant', content: 'FINAL: success' },
@@ -193,6 +208,31 @@ describe('turnLoop', () => {
     expect(step2.result).toMatch(/Success/);
   });
 
+  it('handles Ground Control state locally without dispatching it as a browser tool', async () => {
+    shared.chatMock.mockResolvedValueOnce({
+      message: {
+        role: 'assistant', content: '', tool_calls: [{
+          id: 'tool_gc', type: 'function', function: {
+            name: 'set_ground_control_state',
+            arguments: JSON.stringify({ app: { baseUrl: 'https://example.com' }, session: { loggedIn: false } }),
+          },
+        }],
+      },
+      usage: { total_tokens: 4 },
+    });
+    shared.chatMock.mockResolvedValueOnce({
+      message: { role: 'assistant', content: 'FINAL: done' }, usage: { total_tokens: 2 },
+    });
+
+    const ctx = { steps: [], missionName: 'demo' };
+    const res = await turnLoop(browser, baseMessages(), 2, 0, 0, {}, ctx);
+
+    expect(res.success).toBe(true);
+    expect(ctx.groundControl.app.baseUrl).toBe('https://example.com');
+    expect(ctx.groundControl.session.loggedIn).toBe(false);
+    expect(res.steps[0].events).toContain('[tool ] ← set_ground_control_state result: ✅ Success');
+  });
+
   it('retries a turn on tool error before marking issues', async () => {
     // First turn: tool call → click_text will throw
     shared.chatMock.mockResolvedValueOnce({

package/tests/llmTests/anthropicProvider.test.js

@@ -0,0 +1,41 @@
+import { describe, it, expect, vi } from 'vitest';
+
+const { create } = vi.hoisted(() => ({ create: vi.fn() }));
+vi.mock('@anthropic-ai/sdk', () => ({
+  default: class Anthropic { constructor(options) { this.options = options; this.messages = { create }; } },
+}));
+
+import { AnthropicProvider, fromAnthropicMessage, toAnthropicRequest } from '../../llm/anthropic/anthropicProvider.js';
+
+describe('AnthropicProvider', () => {
+  it('requires an API key', () => {
+    expect(() => new AnthropicProvider()).toThrow(/ANTHROPIC_API_KEY/);
+  });
+
+  it('converts system, images, tool calls, and tool results', () => {
+    const request = toAnthropicRequest([
+      { role: 'system', content: 'Be precise' },
+      { role: 'user', content: [{ type: 'text', text: 'Inspect' }, { type: 'image', mimeType: 'image/png', data: Buffer.from('png') }] },
+      { role: 'assistant', content: '', tool_calls: [{ id: 'call_1', function: { name: 'click', arguments: '{"selector":"#go"}' } }] },
+      { role: 'tool', tool_call_id: 'call_1', content: 'clicked' },
+    ]);
+    expect(request.system).toEqual([{ type: 'text', text: 'Be precise' }]);
+    expect(request.messages[0].content[1].source).toMatchObject({ type: 'base64', media_type: 'image/png' });
+    expect(request.messages[1].content[0]).toMatchObject({ type: 'tool_use', id: 'call_1', name: 'click' });
+    expect(request.messages[2].content[0]).toMatchObject({ type: 'tool_result', tool_use_id: 'call_1' });
+  });
+
+  it('normalizes Claude output and usage', async () => {
+    create.mockResolvedValueOnce({
+      content: [{ type: 'text', text: 'Working' }, { type: 'tool_use', id: 'toolu_1', name: 'navigate', input: { url: 'https://example.com' } }],
+      usage: { input_tokens: 12, output_tokens: 8 },
+    });
+    const result = await new AnthropicProvider({ apiKey: 'sk-ant-test' }).chat({
+      model: 'claude-sonnet-5', messages: [{ role: 'user', content: 'Go' }],
+      tools: [{ type: 'function', function: { name: 'navigate', description: 'Navigate', parameters: { type: 'object' } } }],
+    });
+    expect(create).toHaveBeenCalledWith(expect.objectContaining({ model: 'claude-sonnet-5', max_tokens: 8192 }));
+    expect(result.message).toEqual(fromAnthropicMessage({ content: [{ type: 'text', text: 'Working' }, { type: 'tool_use', id: 'toolu_1', name: 'navigate', input: { url: 'https://example.com' } }] }));
+    expect(result.usage.total_tokens).toBe(20);
+  });
+});

package/tests/llmTests/llmFactory.test.js

@@ -6,6 +6,7 @@ const { shared } = vi.hoisted(() => ({
   shared: {
     openAICtor: vi.fn(),
     geminiCtor: vi.fn(),
+    anthropicCtor: vi.fn(),
   },
 }));
 
@@ -24,6 +25,12 @@ vi.mock('../../llm/gemini/geminiProvider.js', () => ({
     this.chat = vi.fn(async () => ({ message: { role: 'assistant', content: 'ok' }, usage: {} }));
   },
 }));
+vi.mock('../../llm/anthropic/anthropicProvider.js', () => ({
+  AnthropicProvider: function (opts) {
+    shared.anthropicCtor(opts);
+    this.chat = vi.fn(async () => ({ message: { role: 'assistant', content: 'ok' }, usage: {} }));
+  },
+}));
 
 // Import SUT after mocks
 import { getLLM } from '../../llm/llmFactory.js';
@@ -35,6 +42,7 @@ describe('getLLM', () => {
     process.env = { ...ORIGINAL_ENV };
     shared.openAICtor.mockClear();
     shared.geminiCtor.mockClear();
+    shared.anthropicCtor.mockClear();
   });
 
   afterEach(() => {
@@ -72,7 +80,15 @@ describe('getLLM', () => {
     expect(typeof llm.chat).toBe('function');
   });
 
+  it('creates Anthropic provider and accepts claude as an alias', () => {
+    process.env.ANTHROPIC_API_KEY = 'sk-ant-test';
+    getLLM('claude');
+    expect(shared.anthropicCtor).toHaveBeenCalledWith(
+      expect.objectContaining({ apiKey: 'sk-ant-test' })
+    );
+  });
+
   it('throws on unsupported provider', () => {
-    expect(() => getLLM('anthropic')).toThrow(/Unsupported LLM provider/i);
+    expect(() => getLLM('mistral')).toThrow(/Unsupported LLM provider/i);
   });
 });

package/tests/toolsTests/generateHtmlReport.test.js

@@ -21,11 +21,14 @@ describe('generateHtmlReport', () => {
   it('writes an HTML report containing mission and retry metadata', () => {
     const report = {
       runId: 'run_test',
+      llm: { provider: 'anthropic', model: 'claude-sonnet-5' },
+      cli: { version: '1.7.0' },
       missions: [
         {
           missionName: 'Mission A',
           submissionType: 'mission',
           submissionName: 'Main',
+          file: 'missions/login.mission.ts',
           status: 'passed',
           tags: ['smoke'],
           steps: [
@@ -57,11 +60,16 @@ describe('generateHtmlReport', () => {
     expect(html).toContain('⚠️ Turn Issues');
     expect(html).toContain('data-tag="smoke"');
     expect(html).toContain('data-tags="smoke"');
+    expect(html).toContain('CLI version:</strong> 1.7.0');
+    expect(html).toContain('missions/login.mission.ts');
     expect(html).toContain('data-tag="untagged"');
     expect(html).toContain('id="match-count"');
     expect(html).toContain('No missions match these tags');
     expect(html).toContain('Keep missions that do not match visible');
     expect(html).not.toContain('<div class="pill">LLM:');
+    expect(html).not.toContain('<strong>LLM:</strong>');
+    expect(html).toContain('title="Anthropic"');
+    expect(html).toContain('<strong>claude-sonnet-5</strong>');
   });
 
   it('escapes unsafe text in output', () => {
@@ -95,4 +103,14 @@ describe('generateHtmlReport', () => {
     expect(html).toContain('&lt;script&gt;alert(1)&lt;/script&gt;');
     expect(html).toContain('&lt;b&gt;bold&lt;/b&gt;');
   });
+
+  it('renders the OpenAI logo and model without a textual LLM label', () => {
+    const outPath = path.join(tmpDir, 'openai.html');
+    generateHtmlReport({ runId: 'openai_run', llm: { provider: 'openai', model: 'gpt-5.6' } }, outPath);
+    const html = fs.readFileSync(outPath, 'utf8');
+
+    expect(html).toContain('title="OpenAI"');
+    expect(html).toContain('<strong>gpt-5.6</strong>');
+    expect(html).not.toContain('<strong>LLM:</strong>');
+  });
 });

package/tools/chromeBrowser.js

@@ -26,7 +26,7 @@ import { requestHumanInput } from './humanInput.js';
 import { getMfaCode } from './mfaCode.js';
 
 const FILES_DIR = path.join('missions', 'files');
-const REPORTS_DIR = path.join('missions', 'mission_reports');
+const REPORTS_DIR = process.env.TESTRONAUT_OUTPUT_DIR || path.join('missions', 'mission_reports');
 const FILES_LOG = path.join(REPORTS_DIR, 'files.jsonl');
 
 function ensureDir(p) {
@@ -752,7 +752,11 @@ export class ChromeBrowser {
     }
     if (!chosen) {
       await this._debugClickableSnapshot({ scope }).catch(()=>{});
-      await this.page.screenshot({ path: 'missions/mission_reports/screenshots/click_fail.png' }).catch(()=>{});
+      if (process.env.TESTRONAUT_SCREENSHOTS !== '0') {
+        const failurePath = path.join(REPORTS_DIR, 'screenshots', process.env.TESTRONAUT_RUN_ID || '', 'click_fail.png');
+        ensureDir(path.dirname(failurePath));
+        await this.page.screenshot({ path: failurePath }).catch(()=>{});
+      }
       throw new Error('No clickable locator became ready');
     }
     // do the click (standard → hover+click → JS click)
@@ -1183,13 +1187,19 @@ export class ChromeBrowser {
   }
 
   async screenshot({ name = 'screenshot' }) {
+    if (process.env.TESTRONAUT_SCREENSHOTS === '0') {
+      throw new Error('Screenshots are disabled for this run (--no-screenshots).');
+    }
     const timestamp = new Date().toISOString().replace(/[:.]/g, '-');
-    const screenshotPath = `missions/mission_reports/screenshots/${name}_${timestamp}.png`;
+    const screenshotDir = path.join(REPORTS_DIR, 'screenshots', process.env.TESTRONAUT_RUN_ID || '');
+    const screenshotPath = path.join(screenshotDir, `${name}_${timestamp}.png`);
+    ensureDir(screenshotDir);
   
     await this.page.screenshot({ path: screenshotPath });
   
     // ✅ Return relative path for HTML report
-    return `Screenshot saved at: ./screenshots/${name}_${timestamp}.png`;
+    const reportRelativePath = path.relative(REPORTS_DIR, screenshotPath).split(path.sep).join('/');
+    return `Screenshot saved at: ./${reportRelativePath}`;
   }
   
   async upload_file({ selector, fileName, useChooser = false, timeoutMs = 15000 }) {

package/tools/generateHtmlReport.js

@@ -27,7 +27,7 @@ import { normalizeTags } from '../core/tags.js';
  * @returns {string} absolute path to the written HTML file
  */
 export function generateHtmlReport(report, outputPath) {
-  const { runId, startTime, endTime, missions = [], summary = {}, llm = {} } = report;
+  const { runId, startTime, endTime, missions = [], summary = {}, llm = {}, cli = {} } = report;
   const reportTags = normalizeTags(report.tags ?? missions.flatMap(m => m.submissionType === 'mission' ? (m.tags ?? []) : []));
   const durationSec =
     (startTime && endTime)
@@ -42,6 +42,13 @@ export function generateHtmlReport(report, outputPath) {
   const badge = (status) =>
     status === 'passed' ? '✅ Passed' :
     status === 'failed' ? '❌ Failed' : (status || '—');
+  const providerKey = String(llm.provider ?? '').trim().toLowerCase() === 'claude'
+    ? 'anthropic'
+    : String(llm.provider ?? '').trim().toLowerCase();
+  const provider = {
+    openai: { name: 'OpenAI', logo: '<svg viewBox="0 0 24 24" aria-hidden="true"><path fill="currentColor" d="M9.205 8.658v-2.26c0-.19.072-.333.238-.428l4.543-2.616c.619-.357 1.356-.523 2.117-.523 2.854 0 4.662 2.212 4.662 4.566 0 .167 0 .357-.024.547l-4.71-2.759a.797.797 0 00-.856 0l-5.97 3.473zm10.609 8.8V12.06c0-.333-.143-.57-.429-.737l-5.97-3.473 1.95-1.118a.433.433 0 01.476 0l4.543 2.617c1.309.76 2.189 2.378 2.189 3.948 0 1.808-1.07 3.473-2.76 4.163zM7.802 12.703l-1.95-1.142c-.167-.095-.239-.238-.239-.428V5.899c0-2.545 1.95-4.472 4.591-4.472 1 0 1.927.333 2.712.928L8.23 5.067c-.285.166-.428.404-.428.737v6.898zM12 15.128l-2.795-1.57v-3.33L12 8.658l2.795 1.57v3.33L12 15.128zm1.796 7.23c-1 0-1.927-.332-2.712-.927l4.686-2.712c.285-.166.428-.404.428-.737v-6.898l1.974 1.142c.167.095.238.238.238.428v5.233c0 2.545-1.974 4.472-4.614 4.472zm-5.637-5.303l-4.544-2.617c-1.308-.761-2.188-2.378-2.188-3.948A4.482 4.482 0 014.21 6.327v5.423c0 .333.143.571.428.738l5.947 3.449-1.95 1.118a.432.432 0 01-.476 0zm-.262 3.9c-2.688 0-4.662-2.021-4.662-4.519 0-.19.024-.38.047-.57l4.686 2.71c.286.167.571.167.856 0l5.97-3.448v2.26c0 .19-.07.333-.237.428l-4.543 2.616c-.619.357-1.356.523-2.117.523zm5.899 2.83a5.947 5.947 0 005.827-4.756C22.287 18.339 24 15.84 24 13.296c0-1.665-.713-3.282-1.998-4.448.119-.5.19-.999.19-1.498 0-3.401-2.759-5.947-5.946-5.947-.642 0-1.26.095-1.88.31A5.962 5.962 0 0010.205 0a5.947 5.947 0 00-5.827 4.757C1.713 5.447 0 7.945 0 10.49c0 1.666.713 3.283 1.998 4.448-.119.5-.19 1-.19 1.499 0 3.401 2.759 5.946 5.946 5.946.642 0 1.26-.095 1.88-.309a5.96 5.96 0 004.162 1.713z"/></svg>' },
+    anthropic: { name: 'Anthropic', logo: '<img src="data:image/svg+xml;base64,PHN2ZyByb2xlPSJpbWciIHZpZXdCb3g9IjAgMCAyNCAyNCIgeG1sbnM9Imh0dHA6Ly93d3cudzMub3JnLzIwMDAvc3ZnIj4KICA8dGl0bGU+Q2xhdWRlPC90aXRsZT4KICA8cmVjdCB3aWR0aD0iMjQiIGhlaWdodD0iMjQiIHJ4PSI1IiBmaWxsPSIjRDk3NzU3Ii8+CiAgPHBhdGggZmlsbD0iI2ZmZiIgdHJhbnNmb3JtPSJ0cmFuc2xhdGUoMyAzKSBzY2FsZSguNzUpIiBkPSJNNC43MDkgMTUuOTU1bDQuNzItMi42NDcuMDgtLjIzLS4wOC0uMTI4SDkuMmwtLjc5LS4wNDgtMi42OTgtLjA3My0yLjMzOS0uMDk3LTIuMjY2LS4xMjItLjU3MS0uMTIxTDAgMTEuNzg0bC4wNTUtLjM1Mi40OC0uMzIxLjY4Ni4wNiAxLjUyLjEwMyAyLjI3OC4xNTggMS42NTIuMDk3IDIuNDQ5LjI1NWguMzg5bC4wNTUtLjE1Ny0uMTM0LS4wOTgtLjEwMy0uMDk3LTIuMzU4LTEuNTk2LTIuNTUyLTEuNjg4LTEuMzM2LS45NzItLjcyNC0uNDkxLS4zNjQtLjQ2Mi0uMTU4LTEuMDA4LjY1Ni0uNzIyLjg4MS4wNi4yMjUuMDYxLjg5My42ODYgMS45MDggMS40NzYgMi40OTEgMS44MzMuMzY1LjMwNC4xNDUtLjEwMy4wMTktLjA3My0uMTY0LS4yNzQtMS4zNTUtMi40NDYtMS40NDYtMi40OS0uNjQ0LTEuMDMyLS4xNy0uNjE5YTIuOTcgMi45NyAwIDAxLS4xMDQtLjcyOUw2LjI4My4xMzQgNi42OTYgMGwuOTk2LjEzNC40Mi4zNjQuNjIgMS40MTQgMS4wMDIgMi4yMjkgMS41NTUgMy4wMy40NTYuODk4LjI0My44MzIuMDkxLjI1NWguMTU4VjkuMDFsLjEyOC0xLjcwNi4yMzctMi4wOTUuMjMtMi42OTUuMDgtLjc2LjM3Ni0uOTEuNzQ3LS40OTIuNTg0LjI4LjQ4LjY4NS0uMDY3LjQ0NC0uMjg2IDEuODUxLS41NTkgMi45MDMtLjM2NCAxLjk0MmguMjEybC4yNDMtLjI0Mi45ODUtMS4zMDYgMS42NTItMi4wNjQuNzMtLjgyLjg1LS45MDQuNTQ3LS40MzFoMS4wMzNsLjc2IDEuMTI5LS4zNCAxLjE2Ni0xLjA2NCAxLjM0Ny0uODgxIDEuMTQyLTEuMjY0IDEuNy0uNzkgMS4zNi4wNzMuMTEuMTg4LS4wMiAyLjg1Ni0uNjA2IDEuNTQzLS4yOCAxLjg0MS0uMzE1LjgzMy4zODguMDkxLjM5NS0uMzI4LjgwNy0xLjk2OS40ODYtMi4zMDkuNDYyLTMuNDM5LjgxMy0uMDQyLjAzLjA0OS4wNjEgMS41NDkuMTQ2LjY2Mi4wMzZoMS42MjJsMy4wMi4yMjUuNzkuNTIyLjQ3NC42MzgtLjA3OS40ODUtMS4yMTUuNjItMS42NC0uMzg5LTMuODI5LS45MS0xLjMxMi0uMzI5aC0uMTgydi4xMWwxLjA5MyAxLjA2OCAyLjAwNiAxLjgxIDIuNTA5IDIuMzMuMTI3LjU3OC0uMzIyLjQ1NS0uMzQtLjA0OS0yLjIwNS0xLjY1Ny0uODUxLS43NDctMS45MjYtMS42MmgtLjEyOHYuMTdsLjQ0NC42NDkgMi4zNDUgMy41MjEuMTIyIDEuMDgtLjE3LjM1My0uNjA4LjIxMy0uNjY4LS4xMjItMS4zNzQtMS45MjUtMS40MTUtMi4xNjctMS4xNDMtMS45NDMtLjE0LjA4LS42NzQgNy4yNTQtLjMxNi4zNy0uNzI5LjI4LS42MDctLjQ2MS0uMzIyLS43NDcuMzIyLTEuNDc2LjM4OS0xLjkyNC4zMTUtMS41My4yODYtMS45LjE3LS42MzItLjAxMi0uMDQyLS4xNC4wMTgtMS40MzQgMS45NjctMi4xOCAyLjk0NS0xLjcyNiAxLjg0NS0uNDE0LjE2NC0uNzE3LS4zNy4wNjctLjY2Mi40MDEtLjU4OSAyLjM4OC0zLjAzNiAxLjQ0LTEuODgyLjkzLTEuMDg2LS4wMDYtLjE1OGgtLjA1NUw0LjEzMiAxOC41NmwtMS4xMy4xNDYtLjQ4Ny0uNDU2LjA2MS0uNzQ2LjIzMS0uMjQzIDEuOTA4LTEuMzEyLS4wMDYuMDA2eiIvPgo8L3N2Zz4K" alt="" />' },
+  }[providerKey] ?? { name: llm.provider || 'Unknown provider', logo: '' };
   const tagPalette = [
     ['#60a5fa', 'rgba(96,165,250,.18)'], ['#a78bfa', 'rgba(167,139,250,.18)'],
     ['#34d399', 'rgba(52,211,153,.18)'], ['#f472b6', 'rgba(244,114,182,.18)'],
@@ -137,10 +144,13 @@ export function generateHtmlReport(report, outputPath) {
   const missionGroupBlock = (missionName, subs) => {
     const status = groupStatus(subs);
     const totalSteps = subs.reduce((n, s) => n + (Array.isArray(s.steps) ? s.steps.length : 0), 0);
-    const firstStart = Math.min(...subs.map(s => s.startTime || 0).filter(Boolean));
-    const lastEnd    = Math.max(...subs.map(s => s.endTime || 0).filter(Boolean));
-    const groupDur   = (firstStart && lastEnd) ? ((lastEnd - firstStart) / 1000).toFixed(2) : '—';
+    const startTimes = subs.map(s => s.startTime || 0).filter(Boolean);
+    const endTimes = subs.map(s => s.endTime || 0).filter(Boolean);
+    const firstStart = startTimes.length ? Math.min(...startTimes) : null;
+    const lastEnd = endTimes.length ? Math.max(...endTimes) : null;
+    const groupDur = (firstStart != null && lastEnd != null) ? ((lastEnd - firstStart) / 1000).toFixed(2) : '—';
     const tags = normalizeTags(subs.flatMap(s => s.submissionType === 'mission' ? (s.tags ?? []) : []));
+    const sourceFiles = [...new Set(subs.map(s => s.file).filter(Boolean))];
 
     // pre → mission → post
     const order = { premission: 0, mission: 1, postmission: 2 };
@@ -155,7 +165,7 @@ export function generateHtmlReport(report, outputPath) {
           <span class="name">${esc(missionName)}</span>
           <span class="mission-tags">${tags.map(tag => `<span class="tag-small" style="${tagStyle(tag)}">${esc(tag)}</span>`).join('')}</span>
           <span class="status ${status === 'failed' ? 'bad' : 'ok'}">${badge(status)}</span>
-          <span class="meta">submissions: ${subs.length} • steps: ${totalSteps} • duration: ${groupDur}s</span>
+          <span class="meta">${sourceFiles.length ? `${sourceFiles.map(esc).join(', ')} • ` : ''}submissions: ${subs.length} • steps: ${totalSteps} • duration: ${groupDur}s</span>
           <span class="toolbar">
             <button class="btn-mini toggle" data-scope="mission" aria-label="Expand">▼</button>
           </span>
@@ -236,6 +246,9 @@ export function generateHtmlReport(report, outputPath) {
     .header{ max-width:960px; margin:0 auto 16px; padding:16px 18px; }
     h1{ margin:0 0 6px; font-size:22px; font-weight:800; color:#fff; }
     .run-meta{ color:var(--text-muted); display:grid; gap:4px; }
+    .provider-meta{ display:inline-flex; align-items:center; gap:7px; }
+    .provider-logo{ width:20px; height:20px; color:var(--text); flex:none; }
+    .provider-logo img,.provider-logo svg{ display:block; width:20px; height:20px; border:0; border-radius:0; margin:0; background:transparent; }
 
     /* ===== Chips ===== */
     .summary{ display:flex; flex-wrap:wrap; gap:12px; margin:16px auto 20px; max-width:960px; }
@@ -346,10 +359,12 @@ export function generateHtmlReport(report, outputPath) {
   <div class="header glass">
     <h1>🧑‍🚀 <span class="gradient-text">Testronaut Report</span></h1>
     <div class="run-meta">
+      <div class="provider-meta"><span class="provider-logo" title="${esc(provider.name)}" aria-label="${esc(provider.name)}">${provider.logo}</span><strong>${esc(llm.model ?? '—')}</strong></div>
       <div><strong>Run ID:</strong> ${esc(runId ?? '—')}</div>
       <div><strong>Start:</strong> ${esc(startTime ?? '—')}</div>
-      <div><strong>End:</strong> ${esc(endTime ?? '—')} • <strong>Duration:</strong> ${durationSec}s</div>
-      <div><strong>LLM:</strong> ${esc(llm.provider ?? '—')} • <strong>Model:</strong> ${esc(llm.model ?? '—')}</div>
+      <div><strong>End:</strong> ${esc(endTime ?? '—')}</div>
+      <div><strong>Duration:</strong> ${durationSec}s</div>
+      <div><strong>CLI version:</strong> ${esc(cli.version ?? '—')}</div>
     </div>
   </div>
 

package/tools/mfaCode.js

@@ -9,7 +9,10 @@ export const DEV_API_BASE = 'https://staging.api.testronaut.app';
 
 const MIN_REMAINING_SECONDS = 5;
 const MAX_REMAINING_SECONDS = 25;
-const API_DEBUG_LOG_RELATIVE_PATH = 'missions/mission_reports/api-debug.log';
+const API_DEBUG_LOG_RELATIVE_PATH = path.join(
+  process.env.TESTRONAUT_OUTPUT_DIR || 'missions/mission_reports',
+  'api-debug.log'
+);
 const REDACTED = '<redacted>';
 
 const SENSITIVE_LOG_KEYS = new Set([

package/tools/tokenControl.js

@@ -63,6 +63,10 @@ const DEFAULT_LIMITS = [
   { test: /^gemini-2\.5-flash-8b(-|$)/i, tpm: 300000 },
   { test: /^gemini-2\.5-flash(-|$)/i,    tpm: 300000 },
 
+  // Anthropic limits vary by usage tier; this conservative fallback is only for local pacing.
+  { test: /^claude-(opus|sonnet)-/i,     tpm: 80000 },
+  { test: /^claude-haiku-/i,             tpm: 100000 },
+
   // Ultimate fallback for anything else
   { test: /.*/,                         tpm: 150000 },
 ];
@@ -76,7 +80,7 @@ const warnedModels = new Set();
 /* ---------------- Tokenizer helpers ----------------
  * We prefer tiktoken's per-model encoding when available.
  * For unknown models, we pick a close base:
- *  - o200k_base for modern, long-context families (OpenAI 4o/4.1/5, Gemini 2.5, O*)
+ *  - o200k_base for modern, long-context families (OpenAI, Gemini, Claude, O*)
  *  - cl100k_base as a broad fallback
  */
 /**
@@ -94,10 +98,12 @@ function getTokenizer(model) {
 
     // Treat Gemini 2.5 like modern long-context models
     const isGemini25 = /^gemini-2\.5/.test(m);
+    const isClaude = /^claude-/.test(m);
 
     // OpenAI modern families also map well to o200k_base
     const useO200k =
       isGemini25 ||
+      isClaude ||
       m.startsWith('gpt-5') ||
       m.startsWith('gpt-4o') ||
       m.startsWith('gpt-4.1') ||