[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8,"_9":10},"actionData","errors","root",{"_160":161,"_162":163},"routes/_app.x.$handle.$name",{"_98":97,"_32":45,"_107":88,"_27":108,"_13":14,"_109":-5,"_11":110,"_111":112,"_113":114,"_115":116,"_117":118,"_20":-5,"_119":-5,"_120":-5,"_121":25,"_122":123,"_124":125,"_126":-5,"_127":-5,"_128":129},"routes/_app.x.$handle.$name.branch.$branch.version.$version.diff",{"_11":12,"_13":14,"_15":16,"_17":-5,"_18":-5},"project",{"_43":44,"_32":45,"_38":46,"_47":-5,"_48":25,"_49":50,"_51":50,"_52":53,"_54":-5,"_18":55,"_56":-5,"_57":25,"_58":-5,"_59":-5,"_60":61,"_62":63,"_64":65,"_66":67,"_68":69,"_70":-5,"_71":72,"_73":72,"_74":75,"_76":75,"_77":75,"_78":72,"_79":72,"_80":81,"_82":83,"_84":85,"_86":25},"version",3,"changes",[19],"aiChatInfo","userId",{"_20":21,"_22":23,"_24":25,"_26":25,"_27":28,"_29":28},"file",{"_30":31,"_32":28,"_33":34,"_35":37,"_38":39,"_40":14,"_41":42,"_24":25},"previous",{"_30":31,"_32":28,"_33":34,"_35":36},"deleted",false,"deletedHere","path","main.ts","latestPath","fileId","b7ef2866-2007-11f1-a709-42dde27851f2","name","fileType","http","content","import { StateGraph, Annotation, START, END } from \"npm:@langchain/langgraph@^0.2\";\n\n// ─── Configuration ───────────────────────────────────────────────────────────\nconst ANTHROPIC_URL = \"https://api.z.ai/api/anthropic/v1/messages\";\nconst ANTHROPIC_MODEL = \"anthropic/claude-opus-4-20250514\";\nconst ANTHROPIC_API_KEY = Deno.env.get(\"ANTHROPIC_API_KEY\") ?? \"\";\n\n// ─── LLM Helper ──────────────────────────────────────────────────────────────\nasync function callLLM(systemPrompt: string, userPrompt: string): Promise {\n const res = await fetch(ANTHROPIC_URL, {\n method: \"POST\",\n headers: {\n \"Content-Type\": \"application/json\",\n \"x-api-key\": ANTHROPIC_API_KEY,\n \"anthropic-version\": \"2023-06-01\",\n },\n body: JSON.stringify({\n model: ANTHROPIC_MODEL,\n max_tokens: 4096,\n system: systemPrompt,\n messages: [{ role: \"user\", content: userPrompt }],\n }),\n });\n if (!res.ok) {\n const errText = await res.text();\n throw new Error(`LLM call failed (${res.status}): ${errText}`);\n }\n const data = await res.json();\n return data.content?.[0]?.text ?? \"\";\n}\n\n// ─── State Definition ────────────────────────────────────────────────────────\nconst AgentState = Annotation.Root({\n // Input\n mode: Annotation(),\n input: Annotation(),\n\n // Routing\n detectedMode: Annotation(),\n\n // Node outputs\n analysisResult: Annotation(),\n testCases: Annotation(),\n triageResult: Annotation(),\n strategyResult: Annotation(),\n\n // Final\n finalOutput: Annotation(),\n});\n\n// ─── Node: Auto-Detect Mode ─────────────────────────────────────────────────\nasync function detectMode(state: typeof AgentState.State) {\n if (state.mode !== \"auto\") {\n return { detectedMode: state.mode };\n }\n\n const classification = await callLLM(\n `You are a classifier for mobile QA requests. Respond with exactly one word:\n- \"generate\" if the user wants test cases generated for a feature/screen\n- \"triage\" if the user is sharing a test failure, crash log, stack trace, or bug report\n- \"strategy\" if the user wants test strategy advice, test pyramid planning, or coverage analysis\nRespond with ONLY the single word.`,\n state.input,\n );\n\n const mode = classification.trim().toLowerCase();\n const valid = [\"generate\", \"triage\", \"strategy\"];\n return { detectedMode: valid.includes(mode) ? mode : \"generate\" };\n}\n\n// ─── Node: Analyze Input ─────────────────────────────────────────────────────\nasync function analyzeInput(state: typeof AgentState.State) {\n const result = await callLLM(\n `You are a senior mobile QA engineer. Analyze the user's input and extract:\n1. Platform context (iOS, Android, or both)\n2. App area / feature under test\n3. Key risk areas and edge cases to consider\n4. Any implicit requirements (accessibility, localization, offline, etc.)\nRespond in structured JSON with keys: platform, feature, risks, implicitRequirements`,\n state.input,\n );\n return { analysisResult: result };\n}\n\n// ─── Node: Generate Test Cases ───────────────────────────────────────────────\nasync function generateTestCases(state: typeof AgentState.State) {\n const result = await callLLM(\n `You are an expert mobile QA test engineer. Using the analysis context provided, generate comprehensive test cases.\n\nFor each test case provide:\n- id: TC-XXX\n- title: concise description\n- priority: P0 (critical) / P1 (high) / P2 (medium) / P3 (low)\n- type: functional | UI | performance | accessibility | security | edge-case\n- platforms: iOS | Android | both\n- preconditions: what must be true before executing\n- steps: numbered test steps\n- expected: expected result\n- automatable: yes/no with suggested framework (XCUITest, Espresso, Appium, Detox, Maestro)\n\nCover: happy path, negative cases, boundary values, interruptions (calls, notifications, backgrounding), network conditions, accessibility, and device rotation.\n\nRespond in JSON array format.`,\n `User request: ${state.input}\\n\\nAnalysis context: ${state.analysisResult}`,\n );\n return { testCases: result };\n}\n\n// ─── Node: Triage Failure ────────────────────────────────────────────────────\nasync function triageFailure(state: typeof AgentState.State) {\n const result = await callLLM(\n `You are a senior mobile QA engineer specializing in failure analysis. Analyze the provided test failure / crash log / bug report.\n\nProvide:\n1. failureClassification: UI regression | crash | network error | flaky test | data issue | platform-specific bug | race condition | memory leak\n2. severityAssessment: blocker / critical / major / minor\n3. rootCauseAnalysis: Most likely cause based on the symptoms\n4. platformImpact: iOS-only, Android-only, or cross-platform\n5. suggestedFix: Concrete steps the dev team should take\n6. regressionRisk: What other areas might be affected\n7. testImprovement: How to make the test more robust to prevent false positives\n8. automationNote: If relevant, suggest automation improvements (Appium, Espresso, XCUITest, etc.)\n\nRespond in structured JSON.`,\n `Failure input: ${state.input}\\n\\nAnalysis context: ${state.analysisResult}`,\n );\n return { triageResult: result };\n}\n\n// ─── Node: Strategy Advisor ──────────────────────────────────────────────────\nasync function adviseStrategy(state: typeof AgentState.State) {\n const result = await callLLM(\n `You are a mobile test strategy architect. Based on the app description provided, create a comprehensive mobile test strategy.\n\nInclude:\n1. testPyramid: Ratio of unit / integration / E2E tests with justification\n2. frameworkSelection: Recommended frameworks per layer (XCUITest, Espresso, Appium, Detox, Maestro, EarlGrey) with pros/cons for this specific app\n3. platformCoverageMatrix: What to test on each platform, device tiers, OS versions\n4. riskBasedPrioritization: Which features need most coverage and why\n5. cicdIntegration: How to structure the test pipeline (pre-merge, post-merge, nightly, release)\n6. deviceFarmRecommendation: BrowserStack, AWS Device Farm, Firebase Test Lab, Sauce Labs — which fits best\n7. metricsAndKPIs: What to track (flake rate, coverage %, pass rate, execution time)\n8. shiftLeftRecommendations: What can be caught earlier in the development cycle\n\nRespond in structured JSON.`,\n `App description: ${state.input}\\n\\nAnalysis context: ${state.analysisResult}`,\n );\n return { strategyResult: result };\n}\n\n// ─── Node: Assemble Final Output ─────────────────────────────────────────────\nasync function assembleFinal(state: typeof AgentState.State) {\n const mode = state.detectedMode;\n const modeLabel =\n mode === \"generate\" ? \"Test Case Generation\"\n : mode === \"triage\" ? \"Failure Triage\"\n : \"Test Strategy Advisory\";\n\n const primaryResult =\n mode === \"generate\" ? state.testCases\n : mode === \"triage\" ? state.triageResult\n : state.strategyResult;\n\n return {\n finalOutput: JSON.stringify({\n agent: \"Mobile QA Agent (LangGraph)\",\n mode: modeLabel,\n analysis: safeParseJSON(state.analysisResult),\n result: safeParseJSON(primaryResult),\n }, null, 2),\n };\n}\n\nfunction safeParseJSON(str: string): unknown {\n try {\n const cleaned = str.replace(/^```(?:json)?\\n?/gm, \"\").replace(/```$/gm, \"\").trim();\n return JSON.parse(cleaned);\n } catch {\n return str;\n }\n}\n\n// ─── Routing Function ────────────────────────────────────────────────────────\nfunction routeByMode(state: typeof AgentState.State): string {\n switch (state.detectedMode) {\n case \"triage\": return \"triageFailure\";\n case \"strategy\": return \"adviseStrategy\";\n default: return \"generateTestCases\";\n }\n}\n\n// ─── Build the Graph ─────────────────────────────────────────────────────────\nconst workflow = new StateGraph(AgentState)\n .addNode(\"detectMode\", detectMode)\n .addNode(\"analyzeInput\", analyzeInput)\n .addNode(\"generateTestCases\", generateTestCases)\n .addNode(\"triageFailure\", triageFailure)\n .addNode(\"adviseStrategy\", adviseStrategy)\n .addNode(\"assembleFinal\", assembleFinal)\n .addEdge(START, \"detectMode\")\n .addEdge(\"detectMode\", \"analyzeInput\")\n .addConditionalEdges(\"analyzeInput\", routeByMode, {\n generateTestCases: \"generateTestCases\",\n triageFailure: \"triageFailure\",\n adviseStrategy: \"adviseStrategy\",\n })\n .addEdge(\"generateTestCases\", \"assembleFinal\")\n .addEdge(\"triageFailure\", \"assembleFinal\")\n .addEdge(\"adviseStrategy\", \"assembleFinal\")\n .addEdge(\"assembleFinal\", END);\n\nconst app = workflow.compile();\n\n// ─── HTTP Handler ────────────────────────────────────────────────────────────\nexport default async function handler(req: Request): Promise {\n if (req.method === \"OPTIONS\") {\n return new Response(null, {\n headers: {\n \"Access-Control-Allow-Origin\": \"*\",\n \"Access-Control-Allow-Methods\": \"GET, POST, OPTIONS\",\n \"Access-Control-Allow-Headers\": \"Content-Type\",\n },\n });\n }\n\n const corsHeaders = {\n \"Access-Control-Allow-Origin\": \"*\",\n \"Content-Type\": \"application/json\",\n };\n\n // GET → usage info\n if (req.method === \"GET\") {\n return new Response(JSON.stringify({\n name: \"Mobile QA Agent\",\n description: \"LangGraph-powered mobile QA test automation agent combining test case generation, failure triage, and test strategy advisory.\",\n version: \"1.1.0\",\n poweredBy: \"LangGraph StateGraph + Anthropic Claude\",\n capabilities: [\"generate\", \"triage\", \"strategy\", \"auto\"],\n usage: {\n method: \"POST\",\n contentType: \"application/json\",\n body: {\n mode: \"auto | generate | triage | strategy (default: auto)\",\n input: \"Your feature description, failure log, or app description\",\n },\n examples: [\n {\n mode: \"generate\",\n input: \"Login screen with email/password and biometric authentication on iOS and Android\",\n },\n {\n mode: \"triage\",\n input: \"java.lang.NullPointerException at LoginActivity.onBiometricResult(LoginActivity.java:142) — test was passing yesterday, now fails on Android 14 Pixel 8\",\n },\n {\n mode: \"strategy\",\n input: \"E-commerce app with 50K DAU, checkout flow, product search, user profiles, push notifications. Currently zero automation.\",\n },\n {\n mode: \"auto\",\n input: \"Our login tests keep failing intermittently on Samsung devices after the latest Android update\",\n },\n ],\n },\n graph: {\n nodes: [\"detectMode\", \"analyzeInput\", \"generateTestCases\", \"triageFailure\", \"adviseStrategy\", \"assembleFinal\"],\n flow: \"START → detectMode → analyzeInput → [conditional: generate|triage|strategy] → assembleFinal → END\",\n },\n }, null, 2), { headers: corsHeaders });\n }\n\n // POST → run agent\n if (req.method === \"POST\") {\n const startTime = Date.now();\n try {\n const body = await req.json();\n const { mode = \"auto\", input } = body;\n\n if (!input || typeof input !== \"string\") {\n return new Response(JSON.stringify({ error: \"Missing required field: input (string)\" }), {\n status: 400, headers: corsHeaders,\n });\n }\n\n if (!ANTHROPIC_API_KEY) {\n return new Response(JSON.stringify({ error: \"ANTHROPIC_API_KEY environment variable not set\" }), {\n status: 500, headers: corsHeaders,\n });\n }\n\n const validModes = [\"auto\", \"generate\", \"triage\", \"strategy\"];\n if (!validModes.includes(mode)) {\n return new Response(JSON.stringify({ error: `Invalid mode. Use: ${validModes.join(\", \")}` }), {\n status: 400, headers: corsHeaders,\n });\n }\n\n const result = await app.invoke({\n mode,\n input,\n detectedMode: \"\",\n analysisResult: \"\",\n testCases: \"\",\n triageResult: \"\",\n strategyResult: \"\",\n finalOutput: \"\",\n });\n\n const elapsed = Date.now() - startTime;\n const parsed = safeParseJSON(result.finalOutput);\n const output = typeof parsed === \"object\" && parsed !== null\n ? { ...parsed as Record, meta: { executionTimeMs: elapsed, timestamp: new Date().toISOString() } }\n : parsed;\n\n return new Response(JSON.stringify(output, null, 2), { headers: corsHeaders });\n } catch (err) {\n const elapsed = Date.now() - startTime;\n const message = err instanceof Error ? err.message : \"Unknown error\";\n return new Response(JSON.stringify({ error: message, meta: { executionTimeMs: elapsed } }), {\n status: 500, headers: corsHeaders,\n });\n }\n }\n\n return new Response(JSON.stringify({ error: \"Method not allowed\" }), {\n status: 405, headers: corsHeaders,\n });\n}\n","import { StateGraph, Annotation, START, END } from \"npm:@langchain/langgraph@^0.2\";\n\n// ─── Configuration ───────────────────────────────────────────────────────────\nconst ANTHROPIC_URL = \"https://api.z.ai/api/anthropic/v1/messages\";\nconst ANTHROPIC_MODEL = \"anthropic/claude-opus-4-20250514\";\nconst ANTHROPIC_API_KEY = Deno.env.get(\"ANTHROPIC_API_KEY\") ?? \"\";\nconst X_API_KEY = Deno.env.get(\"X_API_KEY\") ?? \"\";\n\n// ─── LLM Helper ──────────────────────────────────────────────────────────────\nasync function callLLM(systemPrompt: string, userPrompt: string): Promise {\n const res = await fetch(ANTHROPIC_URL, {\n method: \"POST\",\n headers: {\n \"Content-Type\": \"application/json\",\n \"x-api-key\": ANTHROPIC_API_KEY,\n \"anthropic-version\": \"2023-06-01\",\n },\n body: JSON.stringify({\n model: ANTHROPIC_MODEL,\n max_tokens: 4096,\n system: systemPrompt,\n messages: [{ role: \"user\", content: userPrompt }],\n }),\n });\n if (!res.ok) {\n const errText = await res.text();\n throw new Error(`LLM call failed (${res.status}): ${errText}`);\n }\n const data = await res.json();\n return data.content?.[0]?.text ?? \"\";\n}\n\n// ─── Auth Helper ─────────────────────────────────────────────────────────────\nfunction authenticateRequest(req: Request): boolean {\n if (!X_API_KEY) return true; // If no key is configured, allow all (dev mode)\n const provided = req.headers.get(\"X-API-Key\") ?? req.headers.get(\"x-api-key\") ?? \"\";\n return provided === X_API_KEY;\n}\n\n// ─── State Definition ────────────────────────────────────────────────────────\nconst AgentState = Annotation.Root({\n mode: Annotation(),\n input: Annotation(),\n detectedMode: Annotation(),\n analysisResult: Annotation(),\n testCases: Annotation(),\n triageResult: Annotation(),\n strategyResult: Annotation(),\n finalOutput: Annotation(),\n});\n\n// ─── Node: Auto-Detect Mode ─────────────────────────────────────────────────\nasync function detectMode(state: typeof AgentState.State) {\n if (state.mode !== \"auto\") {\n return { detectedMode: state.mode };\n }\n\n const classification = await callLLM(\n `You are a classifier for mobile QA requests. Respond with exactly one word:\n- \"generate\" if the user wants test cases generated for a feature/screen\n- \"triage\" if the user is sharing a test failure, crash log, stack trace, or bug report\n- \"strategy\" if the user wants test strategy advice, test pyramid planning, or coverage analysis\nRespond with ONLY the single word.`,\n state.input,\n );\n\n const mode = classification.trim().toLowerCase();\n const valid = [\"generate\", \"triage\", \"strategy\"];\n return { detectedMode: valid.includes(mode) ? mode : \"generate\" };\n}\n\n// ─── Node: Analyze Input ─────────────────────────────────────────────────────\nasync function analyzeInput(state: typeof AgentState.State) {\n const result = await callLLM(\n `You are a senior mobile QA engineer. Analyze the user's input and extract:\n1. Platform context (iOS, Android, or both)\n2. App area / feature under test\n3. Key risk areas and edge cases to consider\n4. Any implicit requirements (accessibility, localization, offline, etc.)\nRespond in structured JSON with keys: platform, feature, risks, implicitRequirements`,\n state.input,\n );\n return { analysisResult: result };\n}\n\n// ─── Node: Generate Test Cases ───────────────────────────────────────────────\nasync function generateTestCases(state: typeof AgentState.State) {\n const result = await callLLM(\n `You are an expert mobile QA test engineer. Using the analysis context provided, generate comprehensive test cases.\n\nFor each test case provide:\n- id: TC-XXX\n- title: concise description\n- priority: P0 (critical) / P1 (high) / P2 (medium) / P3 (low)\n- type: functional | UI | performance | accessibility | security | edge-case\n- platforms: iOS | Android | both\n- preconditions: what must be true before executing\n- steps: numbered test steps\n- expected: expected result\n- automatable: yes/no with suggested framework (XCUITest, Espresso, Appium, Detox, Maestro)\n\nCover: happy path, negative cases, boundary values, interruptions (calls, notifications, backgrounding), network conditions, accessibility, and device rotation.\n\nRespond in JSON array format.`,\n `User request: ${state.input}\\n\\nAnalysis context: ${state.analysisResult}`,\n );\n return { testCases: result };\n}\n\n// ─── Node: Triage Failure ────────────────────────────────────────────────────\nasync function triageFailure(state: typeof AgentState.State) {\n const result = await callLLM(\n `You are a senior mobile QA engineer specializing in failure analysis. Analyze the provided test failure / crash log / bug report.\n\nProvide:\n1. failureClassification: UI regression | crash | network error | flaky test | data issue | platform-specific bug | race condition | memory leak\n2. severityAssessment: blocker / critical / major / minor\n3. rootCauseAnalysis: Most likely cause based on the symptoms\n4. platformImpact: iOS-only, Android-only, or cross-platform\n5. suggestedFix: Concrete steps the dev team should take\n6. regressionRisk: What other areas might be affected\n7. testImprovement: How to make the test more robust to prevent false positives\n8. automationNote: If relevant, suggest automation improvements (Appium, Espresso, XCUITest, etc.)\n\nRespond in structured JSON.`,\n `Failure input: ${state.input}\\n\\nAnalysis context: ${state.analysisResult}`,\n );\n return { triageResult: result };\n}\n\n// ─── Node: Strategy Advisor ──────────────────────────────────────────────────\nasync function adviseStrategy(state: typeof AgentState.State) {\n const result = await callLLM(\n `You are a mobile test strategy architect. Based on the app description provided, create a comprehensive mobile test strategy.\n\nInclude:\n1. testPyramid: Ratio of unit / integration / E2E tests with justification\n2. frameworkSelection: Recommended frameworks per layer (XCUITest, Espresso, Appium, Detox, Maestro, EarlGrey) with pros/cons for this specific app\n3. platformCoverageMatrix: What to test on each platform, device tiers, OS versions\n4. riskBasedPrioritization: Which features need most coverage and why\n5. cicdIntegration: How to structure the test pipeline (pre-merge, post-merge, nightly, release)\n6. deviceFarmRecommendation: BrowserStack, AWS Device Farm, Firebase Test Lab, Sauce Labs — which fits best\n7. metricsAndKPIs: What to track (flake rate, coverage %, pass rate, execution time)\n8. shiftLeftRecommendations: What can be caught earlier in the development cycle\n\nRespond in structured JSON.`,\n `App description: ${state.input}\\n\\nAnalysis context: ${state.analysisResult}`,\n );\n return { strategyResult: result };\n}\n\n// ─── Node: Assemble Final Output ─────────────────────────────────────────────\nasync function assembleFinal(state: typeof AgentState.State) {\n const mode = state.detectedMode;\n const modeLabel =\n mode === \"generate\" ? \"Test Case Generation\"\n : mode === \"triage\" ? \"Failure Triage\"\n : \"Test Strategy Advisory\";\n\n const primaryResult =\n mode === \"generate\" ? state.testCases\n : mode === \"triage\" ? state.triageResult\n : state.strategyResult;\n\n return {\n finalOutput: JSON.stringify({\n agent: \"Mobile QA Agent (LangGraph)\",\n mode: modeLabel,\n analysis: safeParseJSON(state.analysisResult),\n result: safeParseJSON(primaryResult),\n }, null, 2),\n };\n}\n\nfunction safeParseJSON(str: string): unknown {\n try {\n const cleaned = str.replace(/^```(?:json)?\\n?/gm, \"\").replace(/```$/gm, \"\").trim();\n return JSON.parse(cleaned);\n } catch {\n return str;\n }\n}\n\n// ─── Routing Function ────────────────────────────────────────────────────────\nfunction routeByMode(state: typeof AgentState.State): string {\n switch (state.detectedMode) {\n case \"triage\": return \"triageFailure\";\n case \"strategy\": return \"adviseStrategy\";\n default: return \"generateTestCases\";\n }\n}\n\n// ─── Build the Graph ─────────────────────────────────────────────────────────\nconst workflow = new StateGraph(AgentState)\n .addNode(\"detectMode\", detectMode)\n .addNode(\"analyzeInput\", analyzeInput)\n .addNode(\"generateTestCases\", generateTestCases)\n .addNode(\"triageFailure\", triageFailure)\n .addNode(\"adviseStrategy\", adviseStrategy)\n .addNode(\"assembleFinal\", assembleFinal)\n .addEdge(START, \"detectMode\")\n .addEdge(\"detectMode\", \"analyzeInput\")\n .addConditionalEdges(\"analyzeInput\", routeByMode, {\n generateTestCases: \"generateTestCases\",\n triageFailure: \"triageFailure\",\n adviseStrategy: \"adviseStrategy\",\n })\n .addEdge(\"generateTestCases\", \"assembleFinal\")\n .addEdge(\"triageFailure\", \"assembleFinal\")\n .addEdge(\"adviseStrategy\", \"assembleFinal\")\n .addEdge(\"assembleFinal\", END);\n\nconst app = workflow.compile();\n\n// ─── HTTP Handler ────────────────────────────────────────────────────────────\nexport default async function handler(req: Request): Promise {\n if (req.method === \"OPTIONS\") {\n return new Response(null, {\n headers: {\n \"Access-Control-Allow-Origin\": \"*\",\n \"Access-Control-Allow-Methods\": \"GET, POST, OPTIONS\",\n \"Access-Control-Allow-Headers\": \"Content-Type, X-API-Key\",\n },\n });\n }\n\n const corsHeaders = {\n \"Access-Control-Allow-Origin\": \"*\",\n \"Content-Type\": \"application/json\",\n };\n\n // GET → usage info (public, no auth required)\n if (req.method === \"GET\") {\n return new Response(JSON.stringify({\n name: \"Mobile QA Agent\",\n description: \"LangGraph-powered mobile QA test automation agent combining test case generation, failure triage, and test strategy advisory.\",\n version: \"1.2.0\",\n poweredBy: \"LangGraph StateGraph + Anthropic Claude\",\n authentication: {\n type: \"API Key\",\n header: \"X-API-Key\",\n note: \"Required for POST requests. GET is public.\",\n },\n capabilities: [\"generate\", \"triage\", \"strategy\", \"auto\"],\n usage: {\n method: \"POST\",\n contentType: \"application/json\",\n headers: { \"X-API-Key\": \"\" },\n body: {\n mode: \"auto | generate | triage | strategy (default: auto)\",\n input: \"Your feature description, failure log, or app description\",\n },\n examples: [\n {\n mode: \"generate\",\n input: \"Login screen with email/password and biometric authentication on iOS and Android\",\n },\n {\n mode: \"triage\",\n input: \"java.lang.NullPointerException at LoginActivity.onBiometricResult(LoginActivity.java:142) — test was passing yesterday, now fails on Android 14 Pixel 8\",\n },\n {\n mode: \"strategy\",\n input: \"E-commerce app with 50K DAU, checkout flow, product search, user profiles, push notifications. Currently zero automation.\",\n },\n {\n mode: \"auto\",\n input: \"Our login tests keep failing intermittently on Samsung devices after the latest Android update\",\n },\n ],\n },\n graph: {\n nodes: [\"detectMode\", \"analyzeInput\", \"generateTestCases\", \"triageFailure\", \"adviseStrategy\", \"assembleFinal\"],\n flow: \"START → detectMode → analyzeInput → [conditional: generate|triage|strategy] → assembleFinal → END\",\n },\n }, null, 2), { headers: corsHeaders });\n }\n\n // POST → run agent (requires X-API-Key)\n if (req.method === \"POST\") {\n // ── Auth gate ──\n if (!authenticateRequest(req)) {\n return new Response(JSON.stringify({\n error: \"Unauthorized\",\n message: \"Missing or invalid X-API-Key header. This endpoint requires authentication.\",\n }), { status: 401, headers: corsHeaders });\n }\n\n const startTime = Date.now();\n try {\n const body = await req.json();\n const { mode = \"auto\", input } = body;\n\n if (!input || typeof input !== \"string\") {\n return new Response(JSON.stringify({ error: \"Missing required field: input (string)\" }), {\n status: 400, headers: corsHeaders,\n });\n }\n\n if (!ANTHROPIC_API_KEY) {\n return new Response(JSON.stringify({ error: \"ANTHROPIC_API_KEY environment variable not set\" }), {\n status: 500, headers: corsHeaders,\n });\n }\n\n const validModes = [\"auto\", \"generate\", \"triage\", \"strategy\"];\n if (!validModes.includes(mode)) {\n return new Response(JSON.stringify({ error: `Invalid mode. Use: ${validModes.join(\", \")}` }), {\n status: 400, headers: corsHeaders,\n });\n }\n\n const result = await app.invoke({\n mode,\n input,\n detectedMode: \"\",\n analysisResult: \"\",\n testCases: \"\",\n triageResult: \"\",\n strategyResult: \"\",\n finalOutput: \"\",\n });\n\n const elapsed = Date.now() - startTime;\n const parsed = safeParseJSON(result.finalOutput);\n const output = typeof parsed === \"object\" && parsed !== null\n ? { ...parsed as Record, meta: { executionTimeMs: elapsed, timestamp: new Date().toISOString() } }\n : parsed;\n\n return new Response(JSON.stringify(output, null, 2), { headers: corsHeaders });\n } catch (err) {\n const elapsed = Date.now() - startTime;\n const message = err instanceof Error ? err.message : \"Unknown error\";\n return new Response(JSON.stringify({ error: message, meta: { executionTimeMs: elapsed } }), {\n status: 500, headers: corsHeaders,\n });\n }\n }\n\n return new Response(JSON.stringify({ error: \"Method not allowed\" }), {\n status: 405, headers: corsHeaders,\n });\n}\n","createdAt",["D",1773537768533],"branchVersion","branchId","9653515a-2007-11f1-a5c4-42dde27851f2","id","9652ee22-2007-11f1-a5c4-42dde27851f2","mobile-qa-agent",["D",1773535238405],"deletedAt","legacyEnvVars","privacy","public","httpPrivacy","description","LangGraph mobile QA agent: test gen, triage, strategy","imageUrl","f404441a-b070-431b-b152-17a2208cfa70","pinnedFileId","pinPreview","pinMigrated","screenshotVersion","likeCount",0,"orgId","018b67b5-3681-41ae-8e5a-cdcf7cd1218d","tursoName","9652ee22200711f1a5c442dde27851f2","tursoHostname","9652ee22200711f1a5c442dde27851f2-val-town.aws-us-east-2.turso.io","tursoMeasuredAt",["D",1789528852539],"pinned","userPerm","r","valPerm","blobPerm","rw","emailPerm","sqlitePerm","projectPerm","telemetryPerm","user",{"_98":97,"_106":102},"org",{"_98":97,"_99":100,"_101":102,"_18":55,"_103":104,"_105":25},"branch",{"_13":87,"_32":88,"_43":42,"_89":44,"_90":-5,"_91":-5,"_92":-5,"_93":-5,"_94":-5,"_95":45,"_96":97},"isOwned",21,"main","projectId","forkedBranchId","forkedBranchVersion","forkedProjectPrivacy","forkedProjectId","forkedProjectOrgId","projectName","userHandle","khawjaahmad","handle","displayName","ahmad","avatar","https://imagedelivery.net/iHX6Ovru0O7AjmyT5yZRoA/4ec015ce-90e5-4eef-7eaf-21e3da8cec00/square512?exp=1789577999&sig=12e186c2f8c7642d2cb3c4df3da9a5f273c768bd126228b99d4ae8207460ff9d","tier","free","canRestrictVals","profileImageUrl","branchName","","adoptOrgId",{"_43":44,"_32":45,"_38":155,"_47":-5,"_48":25,"_49":50,"_51":50,"_52":53,"_54":-5,"_18":55,"_56":-5,"_57":25,"_58":-5,"_59":-5,"_60":61,"_62":63,"_64":65,"_66":67,"_68":156,"_70":-5,"_71":72,"_73":72,"_74":75,"_76":75,"_77":75,"_78":72,"_79":72,"_80":157,"_82":158,"_84":159,"_86":25},"tree",[150],"files",[150],"counts",{"_139":61,"_140":141,"_142":61,"_143":144,"_145":146},"contents",[130],"upstreamProject","upstreamBranch","emptyEnvVars","flags",[],"hasHttpFiles",true,"grants","orgs","customDomains",{},{"_43":131,"_32":28,"_30":31,"_132":-5,"_41":42,"_40":14,"_133":25,"_38":134,"_13":135,"_33":34,"_35":37,"_24":25,"_136":137,"_27":138},"7a67f148-200d-11f1-8dcd-42dde27851f2","parentId","latest",["D",1773537768533],2,"label","mobile-qa",[28],"pulls","branches",1,"forks","envVars",{"_147":135,"_149":61},"connections",{"_147":61,"_148":61},"count","unfilledCount","emptyCount",{"_43":131,"_32":28,"_30":31,"_132":-5,"_41":42,"_40":14,"_133":25,"_38":151,"_13":135,"_33":34,"_35":37,"_24":25,"_136":137,"_152":153,"_27":154},["D",1773537768533],"children",[],[28],["D",1773535238405],["D",1789528852539],{"_98":97,"_106":102},{"_98":97,"_99":100,"_101":102,"_18":55,"_103":104,"_105":25},{"_13":87,"_32":88,"_43":42,"_89":44,"_90":-5,"_91":-5,"_92":-5,"_93":-5,"_94":-5,"_95":45,"_96":97},"ctx",{"_80":-5,"_127":193,"_194":195,"_196":197,"_198":25,"_199":25,"_200":25,"_201":125,"_202":-5,"_203":-7,"_204":-7,"_205":61,"_206":25,"_207":25,"_208":-5,"_209":25,"_210":25,"_211":25,"_212":-5},"ENV",{"_164":165,"_166":167,"_168":88,"_169":170,"_171":172,"_173":174,"_175":176,"_177":178,"_179":180,"_181":182,"_183":184,"_185":186,"_187":188,"_189":190,"_191":192},"REACT_APP_SENTRY_DSN","https://586020d86f05489db63a554690acf1e9@o1398359.ingest.sentry.io/6724499","RENDER_GIT_COMMIT","e87eb04e707cf64e7e532c8e8ec395f209e17f15","RENDER_GIT_BRANCH","POSTHOG_API_TOKEN","phc_nmoyuAkStEZuUp69rYlg1DZgrIt19Gx0Y4Hwf8tGgJP","POSTHOG_API_HOST","https://wart.val.town","WEB_API_TEMPLATE","https://{val}.web.val.run","APP_URL","https://www.val.town","MODULE_URL","https://esm.town","VANITY_WEB_API_TEMPLATE","https://{label}.val.run/","DENOLS_URL","https://lsp.val.town","NODE_ENV","production","CLOUDFLARE_TURNSTILE_SITE_KEY","0x4AAAAAACtaIZJH7s9rBMZ6","GOOGLE_CLIENT_ID","875989696899-jtsqkhlbdv5jrlsk8631p7ipjg3c6o62.apps.googleusercontent.com","STRIPE_PUBLISHABLE_KEY","pk_live_51N2zvWEpZTTLipIkVEwXrUG9AT0yeOOEYxszarZT1jpsZV58Ly6wlg3SltZpHN26CUalOBmUSY7adQRQiSkCZz2500qSXNkf8Y","API_URL","https://api.val.town",[],"preferencesKeybindings","default","preferencesTheme","light","preferencesFixers","preferencesCopilot","preferencesSelectionAi","preferencesCloseBrackets","preferencesChatPanelState","preferencesDprint","preferencesDenofmt","notificationCount","flagDenols","isBusinessOrgMember","businessOrgHandle","isBusinessOrgAdmin","isHigherTownieTier","navCollapsed","currentOrgId"]

Diff · v3