@pauly4010/evalai-sdk 1.3.0 → 1.4.1

package/dist/cli/check.js

@@ -0,0 +1,215 @@
+#!/usr/bin/env node
+"use strict";
+/**
+ * evalai check — CI/CD evaluation gate
+ *
+ * Usage:
+ *   evalai check --minScore 92 --evaluationId 42
+ *   evalai check --minScore 90 --maxDrop 5 --evaluationId 42
+ *   evalai check --policy HIPAA --evaluationId 42
+ *   evalai check --baseline published --evaluationId 42
+ *
+ * Flags:
+ *   --minScore <n>       Fail if quality score < n (0-100)
+ *   --maxDrop <n>        Fail if score dropped > n points from baseline
+ *   --minN <n>           Fail if total test cases < n (low sample size)
+ *   --allowWeakEvidence  If false (default), fail when evidenceLevel is 'weak'
+ *   --policy <name>      Enforce a compliance policy (e.g. HIPAA, SOC2, GDPR)
+ *   --baseline <mode>    Baseline comparison mode: "published" (default), "previous", or "production"
+ *   --evaluationId <id>  Required. The evaluation to gate on.
+ *   --baseUrl <url>      API base URL (default: EVALAI_BASE_URL or http://localhost:3000)
+ *   --apiKey <key>       API key (default: EVALAI_API_KEY env var)
+ *
+ * Exit codes:
+ *   0  — Gate passed
+ *   1  — Gate failed: score below threshold
+ *   2  — Gate failed: regression exceeded maxDrop
+ *   3  — Gate failed: policy violation
+ *   4  — API error / network failure
+ *   5  — Invalid arguments
+ *   6  — Gate failed: total test cases < minN
+ *   7  — Gate failed: weak evidence (evidenceLevel === 'weak')
+ *
+ * Environment:
+ *   EVALAI_BASE_URL  — API base URL (default: http://localhost:3000)
+ *   EVALAI_API_KEY   — API key for authentication
+ */
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.EXIT = void 0;
+exports.parseArgs = parseArgs;
+exports.runCheck = runCheck;
+// Standardized exit codes
+exports.EXIT = {
+    PASS: 0,
+    SCORE_BELOW: 1,
+    REGRESSION: 2,
+    POLICY_VIOLATION: 3,
+    API_ERROR: 4,
+    BAD_ARGS: 5,
+    LOW_N: 6,
+    WEAK_EVIDENCE: 7,
+};
+function parseArgs(argv) {
+    const args = {};
+    for (let i = 0; i < argv.length; i++) {
+        const arg = argv[i];
+        if (arg.startsWith('--')) {
+            const key = arg.slice(2);
+            const next = argv[i + 1];
+            if (next !== undefined && !next.startsWith('--')) {
+                args[key] = next;
+                i++;
+            }
+            else {
+                args[key] = 'true'; // bare flag
+            }
+        }
+    }
+    const baseUrl = args.baseUrl || process.env.EVALAI_BASE_URL || 'http://localhost:3000';
+    const apiKey = args.apiKey || process.env.EVALAI_API_KEY || '';
+    const minScore = parseInt(args.minScore || '0');
+    const maxDrop = args.maxDrop ? parseInt(args.maxDrop) : undefined;
+    const minN = args.minN ? parseInt(args.minN) : undefined;
+    const allowWeakEvidence = args.allowWeakEvidence === 'true' || args.allowWeakEvidence === '1';
+    const evaluationId = args.evaluationId || '';
+    const policy = args.policy || undefined;
+    const baseline = (args.baseline === 'previous'
+        ? 'previous'
+        : args.baseline === 'production'
+            ? 'production'
+            : 'published');
+    if (!apiKey) {
+        console.error('Error: --apiKey or EVALAI_API_KEY is required');
+        process.exit(exports.EXIT.BAD_ARGS);
+    }
+    if (!evaluationId) {
+        console.error('Error: --evaluationId is required');
+        process.exit(exports.EXIT.BAD_ARGS);
+    }
+    if (isNaN(minScore) || minScore < 0 || minScore > 100) {
+        console.error('Error: --minScore must be 0-100');
+        process.exit(exports.EXIT.BAD_ARGS);
+    }
+    if (minN !== undefined && (isNaN(minN) || minN < 1)) {
+        console.error('Error: --minN must be a positive number');
+        process.exit(exports.EXIT.BAD_ARGS);
+    }
+    return { baseUrl, apiKey, minScore, maxDrop, minN, allowWeakEvidence, evaluationId, policy, baseline };
+}
+async function runCheck(args) {
+    const headers = { Authorization: `Bearer ${args.apiKey}` };
+    // ── 1. Fetch latest quality score ──
+    const scoreUrl = `${args.baseUrl}/api/quality?evaluationId=${args.evaluationId}&action=latest&baseline=${args.baseline}`;
+    let scoreRes;
+    try {
+        scoreRes = await fetch(scoreUrl, { headers });
+    }
+    catch (err) {
+        console.error(`EvalAI gate ERROR: Network failure — ${err.message}`);
+        return exports.EXIT.API_ERROR;
+    }
+    if (!scoreRes.ok) {
+        const body = await scoreRes.text();
+        console.error(`EvalAI gate ERROR: API returned ${scoreRes.status} — ${body}`);
+        return exports.EXIT.API_ERROR;
+    }
+    const data = (await scoreRes.json());
+    const score = data?.score ?? 0;
+    const total = data?.total ?? null;
+    const evidenceLevel = data?.evidenceLevel ?? null;
+    const baselineScore = data?.baselineScore ?? null;
+    const regressionDelta = data?.regressionDelta ?? null;
+    const baselineMissing = data?.baselineMissing === true;
+    const breakdown = data?.breakdown ?? {};
+    // ── Gate: baseline missing (when baseline comparison requested) ──
+    if (baselineMissing && (args.baseline !== 'published' || args.maxDrop !== undefined)) {
+        console.error(`\n✗ FAILED: baseline (${args.baseline}) not found. ` +
+            `Ensure a baseline run exists (e.g. published run, previous run, or prod-tagged run).`);
+        return exports.EXIT.API_ERROR;
+    }
+    // ── Gate: minN (low sample size) ──
+    if (args.minN !== undefined && total !== null && total < args.minN) {
+        console.error(`\n✗ FAILED: total test cases (${total}) < minN (${args.minN})`);
+        return exports.EXIT.LOW_N;
+    }
+    // ── Gate: allowWeakEvidence ──
+    if (!args.allowWeakEvidence && evidenceLevel === 'weak') {
+        console.error(`\n✗ FAILED: evidence level is 'weak' (use --allowWeakEvidence to permit)`);
+        return exports.EXIT.WEAK_EVIDENCE;
+    }
+    // ── Print summary ──
+    console.log('┌─────────────────────────────────────────┐');
+    console.log(`│  EvalAI Quality Score: ${String(score).padStart(3)}/100            │`);
+    console.log('├─────────────────────────────────────────┤');
+    if (baselineScore !== null) {
+        const delta = regressionDelta ?? 0;
+        const arrow = delta >= 0 ? '▲' : '▼';
+        console.log(`│  Baseline: ${baselineScore}  ${arrow} ${Math.abs(delta)} pts          │`);
+    }
+    if (breakdown) {
+        const pct = (v) => `${Math.round((v ?? 0) * 100)}%`;
+        console.log(`│  Pass: ${pct(breakdown.passRate)}  Safety: ${pct(breakdown.safety)}  Judge: ${pct(breakdown.judge)} │`);
+    }
+    if (data?.flags && data.flags.length > 0) {
+        console.log(`│  Flags: ${data.flags.join(', ').padEnd(30)} │`);
+    }
+    console.log('└─────────────────────────────────────────┘');
+    // ── 2. Gate: minimum score ──
+    if (args.minScore > 0 && score < args.minScore) {
+        console.error(`\n✗ FAILED: score=${score} < minScore=${args.minScore}`);
+        return exports.EXIT.SCORE_BELOW;
+    }
+    // ── 3. Gate: maximum drop from baseline ──
+    if (args.maxDrop !== undefined && regressionDelta !== null && regressionDelta < -(args.maxDrop)) {
+        console.error(`\n✗ FAILED: score dropped ${Math.abs(regressionDelta)} pts from baseline ` +
+            `(max allowed: ${args.maxDrop})`);
+        return exports.EXIT.REGRESSION;
+    }
+    // ── 4. Gate: policy compliance ──
+    if (args.policy) {
+        const policyUrl = `${args.baseUrl}/api/quality?evaluationId=${args.evaluationId}&action=latest`;
+        // Check policy-specific flags
+        const policyFlags = (data?.flags ?? []);
+        // Policy mapping: each policy has a set of required conditions
+        const policyChecks = {
+            HIPAA: { requiredSafetyRate: 0.99, maxFlags: ['SAFETY_RISK'] },
+            SOC2: { requiredSafetyRate: 0.95, maxFlags: ['SAFETY_RISK', 'LOW_PASS_RATE'] },
+            GDPR: { requiredSafetyRate: 0.95, maxFlags: ['SAFETY_RISK'] },
+            PCI_DSS: { requiredSafetyRate: 0.99, maxFlags: ['SAFETY_RISK', 'LOW_PASS_RATE'] },
+            FINRA_4511: { requiredSafetyRate: 0.95, maxFlags: ['SAFETY_RISK'] },
+        };
+        const policyName = args.policy.toUpperCase();
+        const check = policyChecks[policyName];
+        if (!check) {
+            console.error(`\n✗ Unknown policy: ${args.policy}. Available: ${Object.keys(policyChecks).join(', ')}`);
+            return exports.EXIT.BAD_ARGS;
+        }
+        // Check safety rate
+        const safetyRate = breakdown?.safety ?? 0;
+        if (safetyRate < check.requiredSafetyRate) {
+            console.error(`\n✗ POLICY VIOLATION (${policyName}): safety rate ${Math.round(safetyRate * 100)}% < ` +
+                `required ${Math.round(check.requiredSafetyRate * 100)}%`);
+            return exports.EXIT.POLICY_VIOLATION;
+        }
+        // Check for disqualifying flags
+        const violations = policyFlags.filter(f => check.maxFlags.includes(f));
+        if (violations.length > 0) {
+            console.error(`\n✗ POLICY VIOLATION (${policyName}): ${violations.join(', ')}`);
+            return exports.EXIT.POLICY_VIOLATION;
+        }
+        console.log(`\n✓ Policy ${policyName}: COMPLIANT`);
+    }
+    console.log('\n✓ EvalAI gate PASSED');
+    return exports.EXIT.PASS;
+}
+// Main entry point
+const isDirectRun = typeof require !== 'undefined' && require.main === module;
+if (isDirectRun) {
+    const args = parseArgs(process.argv.slice(2));
+    runCheck(args).then((code) => {
+        process.exit(code);
+    }).catch((err) => {
+        console.error(`EvalAI gate ERROR: ${err.message}`);
+        process.exit(exports.EXIT.API_ERROR);
+    });
+}

package/dist/cli/index.d.ts

@@ -1,6 +1,8 @@
 #!/usr/bin/env node
 /**
- * CLI for AI Evaluation Platform SDK
- * Tier 2.6: CLI for Everything
+ * evalai — EvalAI CLI
+ *
+ * Commands:
+ *   evalai check  — CI/CD evaluation gate (see evalai check --help)
  */
 export {};

package/dist/cli/index.js

@@ -1,181 +1,44 @@
 #!/usr/bin/env node
 "use strict";
 /**
- * CLI for AI Evaluation Platform SDK
- * Tier 2.6: CLI for Everything
+ * evalai — EvalAI CLI
+ *
+ * Commands:
+ *   evalai check  — CI/CD evaluation gate (see evalai check --help)
  */
-var __importDefault = (this && this.__importDefault) || function (mod) {
-    return (mod && mod.__esModule) ? mod : { "default": mod };
-};
 Object.defineProperty(exports, "__esModule", { value: true });
-const commander_1 = require("commander");
-const promises_1 = __importDefault(require("fs/promises"));
-const path_1 = __importDefault(require("path"));
-const client_1 = require("../client");
-const export_1 = require("../export");
-const program = new commander_1.Command();
-program
-    .name('evalai')
-    .description('AI Evaluation Platform CLI')
-    .version('1.0.0');
-// Initialize project
-program
-    .command('init')
-    .description('Initialize a new evaluation project')
-    .option('-d, --dir <directory>', 'Project directory', '.')
-    .action(async (options) => {
-    const dir = path_1.default.resolve(options.dir);
-    console.log('🚀 Initializing EvalAI project...');
-    // Create directory structure
-    await promises_1.default.mkdir(path_1.default.join(dir, '.evalai'), { recursive: true });
-    await promises_1.default.mkdir(path_1.default.join(dir, '.evalai', 'snapshots'), { recursive: true });
-    await promises_1.default.mkdir(path_1.default.join(dir, 'evaluations'), { recursive: true });
-    // Create config file
-    const config = {
-        apiKey: process.env.EVALAI_API_KEY || '',
-        projectId: process.env.EVALAI_PROJECT_ID || '',
-        baseUrl: 'http://localhost:3000/api',
-        debug: false,
-        retry: {
-            maxAttempts: 3,
-            backoff: 'exponential'
-        }
-    };
-    await promises_1.default.writeFile(path_1.default.join(dir, 'evalai.config.json'), JSON.stringify(config, null, 2));
-    // Create example evaluation file
-    const exampleEval = `import { AIEvalClient, createTestSuite, expect } from '@pauly4010/evalai-sdk'
-
-const client = AIEvalClient.init()
-
-const suite = createTestSuite('example-evaluation', {
-  cases: [
-    {
-      input: 'What is 2+2?',
-      expected: '4',
-      name: 'simple-math'
-    },
-    {
-      input: 'Explain AI in simple terms',
-      expected: (output) => {
-        expect(output).toContainKeywords(['artificial', 'intelligence'])
-        expect(output).toHaveLength({ min: 50, max: 500 })
-        return true
-      },
-      name: 'ai-explanation'
-    }
-  ]
-})
-
-// Run the test suite
-suite.run().then(results => {
-  console.log('Test Results:', results)
-  console.log(\`Passed: \${results.passed}/\${results.total}\`)
-})
-`;
-    await promises_1.default.writeFile(path_1.default.join(dir, 'evaluations', 'example.ts'), exampleEval);
-    console.log('✅ Project initialized successfully!');
-    console.log('\nNext steps:');
-    console.log('1. Set your API key: export EVALAI_API_KEY=your-key');
-    console.log('2. Set your project ID: export EVALAI_PROJECT_ID=your-project');
-    console.log('3. Run evaluations: npx evalai eval:run');
-});
-// Run evaluations
-program
-    .command('eval:run')
-    .description('Run evaluation tests')
-    .option('-c, --config <path>', 'Config file path', './evalai.config.json')
-    .option('-f, --file <path>', 'Evaluation file to run')
-    .action(async (options) => {
-    console.log('🧪 Running evaluations...');
-    // Load config
-    const configPath = path_1.default.resolve(options.config);
-    let config;
-    try {
-        const configContent = await promises_1.default.readFile(configPath, 'utf-8');
-        config = JSON.parse(configContent);
-    }
-    catch (error) {
-        console.error('❌ Config file not found. Run "evalai init" first.');
-        process.exit(1);
-    }
-    const client = client_1.AIEvalClient.init(config);
-    // If file specified, run that file
-    if (options.file) {
-        console.log(`Running ${options.file}...`);
-        // Dynamic import of evaluation file would go here
-        // This requires compilation step for TS files
-    }
-    else {
-        // Run all evaluations in the evaluations directory
-        console.log('Running all evaluations...');
-    }
-    console.log('✅ Evaluations completed!');
-});
-// List traces
-program
-    .command('traces')
-    .description('List and filter traces')
-    .option('-l, --limit <number>', 'Number of traces to show', '10')
-    .option('--failed', 'Show only failed traces')
-    .option('--slow', 'Show slow traces (>5s)')
-    .action(async (options) => {
-    const configPath = path_1.default.resolve('./evalai.config.json');
-    let config;
-    try {
-        const configContent = await promises_1.default.readFile(configPath, 'utf-8');
-        config = JSON.parse(configContent);
-    }
-    catch (error) {
-        console.error('❌ Config file not found. Run "evalai init" first.');
-        process.exit(1);
-    }
-    const client = client_1.AIEvalClient.init(config);
-    console.log('📊 Fetching traces...');
-    // API call to get traces would go here
-    console.log(`Showing ${options.limit} traces`);
-});
-// Export data
-program
-    .command('export')
-    .description('Export data from EvalAI')
-    .option('-f, --format <format>', 'Export format (json, csv, xlsx)', 'json')
-    .option('-o, --output <path>', 'Output file path', './export')
-    .option('-t, --type <type>', 'Data type (traces, evaluations, all)', 'all')
-    .action(async (options) => {
-    const configPath = path_1.default.resolve('./evalai.config.json');
-    let config;
-    try {
-        const configContent = await promises_1.default.readFile(configPath, 'utf-8');
-        config = JSON.parse(configContent);
-    }
-    catch (error) {
-        console.error('❌ Config file not found. Run "evalai init" first.');
-        process.exit(1);
-    }
-    const client = client_1.AIEvalClient.init(config);
-    console.log(`📥 Exporting data as ${options.format}...`);
-    const data = await (0, export_1.exportData)(client, {
-        format: options.format,
-        includeTraces: true,
-        includeEvaluations: true
+const check_1 = require("./check");
+const argv = process.argv.slice(2);
+const subcommand = argv[0];
+if (subcommand === 'check') {
+    const args = (0, check_1.parseArgs)(argv.slice(1));
+    (0, check_1.runCheck)(args)
+        .then((code) => process.exit(code))
+        .catch((err) => {
+        console.error(`EvalAI ERROR: ${err instanceof Error ? err.message : String(err)}`);
+        process.exit(4);
     });
-    // Save to file
-    const outputPath = path_1.default.resolve(process.cwd(), options.output);
-    await promises_1.default.writeFile(outputPath, JSON.stringify(data, null, 2));
-    console.log(`✅ Data exported to ${outputPath}`);
-});
-// Dev server
-program
-    .command('dev')
-    .description('Start local development server')
-    .option('-p, --port <port>', 'Port number', '3001')
-    .action(async (options) => {
-    console.log(`🚀 Starting development server on port ${options.port}...`);
-    console.log('📊 Dashboard: http://localhost:' + options.port);
-    console.log('🔍 API: http://localhost:' + options.port + '/api');
-    console.log('\nPress Ctrl+C to stop');
-    // This would start an Express server with a simple dashboard
-    // For now, just keep the process running
-    process.stdin.resume();
-});
-program.parse();
+}
+else {
+    console.log(`EvalAI CLI
+
+Usage:
+  evalai check [options]   CI/CD evaluation gate
+
+Options for check:
+  --evaluationId <id>  Required. Evaluation to gate on.
+  --apiKey <key>      API key (or EVALAI_API_KEY env)
+  --minScore <n>      Fail if score < n (0-100)
+  --maxDrop <n>       Fail if score dropped > n from baseline
+  --minN <n>          Fail if total test cases < n
+  --allowWeakEvidence Allow weak evidence level
+  --policy <name>     Enforce policy (HIPAA, SOC2, GDPR, etc.)
+  --baseline <mode>   "published" or "previous"
+  --baseUrl <url>     API base URL
+
+Examples:
+  evalai check --minScore 92 --evaluationId 42 --apiKey $EVALAI_API_KEY
+  evalai check --policy HIPAA --evaluationId 42 --apiKey $EVALAI_API_KEY
+`);
+    process.exit(subcommand === '--help' || subcommand === '-h' ? 0 : 1);
+}

package/dist/client.d.ts

@@ -1,4 +1,4 @@
-import { ClientConfig, Trace, CreateTraceParams, ListTracesParams, Evaluation, CreateEvaluationParams, UpdateEvaluationParams, ListEvaluationsParams, LLMJudgeResult, RunLLMJudgeParams, TestCase, CreateTestCaseParams, EvaluationRun, CreateRunParams, Span, CreateSpanParams, OrganizationLimits, Annotation, CreateAnnotationParams, ListAnnotationsParams, AnnotationTask, CreateAnnotationTaskParams, ListAnnotationTasksParams, AnnotationItem, CreateAnnotationItemParams, ListAnnotationItemsParams, APIKey, APIKeyWithSecret, CreateAPIKeyParams, UpdateAPIKeyParams, ListAPIKeysParams, APIKeyUsage, Webhook, CreateWebhookParams, UpdateWebhookParams, ListWebhooksParams, WebhookDelivery, ListWebhookDeliveriesParams, UsageStats, GetUsageParams, UsageSummary, LLMJudgeConfig, CreateLLMJudgeConfigParams, ListLLMJudgeConfigsParams, ListLLMJudgeResultsParams, LLMJudgeAlignment, GetLLMJudgeAlignmentParams, Organization } from './types';
+import { ClientConfig, Trace, CreateTraceParams, ListTracesParams, Evaluation, CreateEvaluationParams, UpdateEvaluationParams, ListEvaluationsParams, LLMJudgeResult, RunLLMJudgeParams, TestCase, CreateTestCaseParams, EvaluationRun, CreateRunParams, Span, CreateSpanParams, UpdateTraceParams, OrganizationLimits, Annotation, CreateAnnotationParams, ListAnnotationsParams, AnnotationTask, CreateAnnotationTaskParams, ListAnnotationTasksParams, AnnotationItem, CreateAnnotationItemParams, ListAnnotationItemsParams, APIKey, APIKeyWithSecret, CreateAPIKeyParams, UpdateAPIKeyParams, ListAPIKeysParams, APIKeyUsage, Webhook, CreateWebhookParams, UpdateWebhookParams, ListWebhooksParams, WebhookDelivery, ListWebhookDeliveriesParams, UsageStats, GetUsageParams, UsageSummary, LLMJudgeConfig, CreateLLMJudgeConfigParams, ListLLMJudgeConfigsParams, ListLLMJudgeResultsParams, LLMJudgeAlignment, GetLLMJudgeAlignmentParams, Organization } from './types';
 import { Logger } from './logger';
 /**
  * AI Evaluation Platform SDK Client
@@ -126,6 +126,19 @@ declare class TraceAPI {
      * Get a single trace by ID
      */
     get(id: number): Promise<Trace>;
+    /**
+     * Update an existing trace (e.g. set status, duration, metadata on completion)
+     *
+     * @example
+     * ```typescript
+     * await client.traces.update(42, {
+     *   status: 'success',
+     *   durationMs: 1234,
+     *   metadata: { output: 'done' }
+     * });
+     * ```
+     */
+    update<TMetadata = Record<string, any>>(id: number, params: UpdateTraceParams<TMetadata>): Promise<Trace<TMetadata>>;
     /**
      * Create a span for a trace
      */

package/dist/client.js

@@ -76,13 +76,45 @@ class AIEvalClient {
         this.cache = new cache_1.RequestCache(config.cacheSize || 1000);
         // Initialize request batcher if enabled (default: enabled)
         if (config.enableBatching !== false) {
+            const MAX_CONCURRENCY = 5;
             this.batcher = new batch_1.RequestBatcher(async (requests) => {
-                // Batch execution placeholder - will be implemented per API
-                return requests.map(req => ({
-                    id: req.id,
-                    status: 200,
-                    data: null,
-                }));
+                const results = [];
+                const executing = [];
+                for (const req of requests) {
+                    const task = (async () => {
+                        try {
+                            const data = await this.request(req.endpoint, {
+                                method: req.method,
+                                body: req.body ? JSON.stringify(req.body) : undefined,
+                                headers: req.headers,
+                            });
+                            results.push({ id: req.id, status: 200, data });
+                        }
+                        catch (err) {
+                            results.push({
+                                id: req.id,
+                                status: err?.statusCode || 500,
+                                data: null,
+                                error: err?.message || 'Unknown error',
+                            });
+                        }
+                    })();
+                    executing.push(task);
+                    if (executing.length >= MAX_CONCURRENCY) {
+                        await Promise.race(executing);
+                        // Remove settled promises
+                        for (let i = executing.length - 1; i >= 0; i--) {
+                            const settled = await Promise.race([
+                                executing[i].then(() => true),
+                                Promise.resolve(false),
+                            ]);
+                            if (settled)
+                                executing.splice(i, 1);
+                        }
+                    }
+                }
+                await Promise.allSettled(executing);
+                return results;
             }, {
                 maxBatchSize: config.batchSize || 10,
                 batchDelay: config.batchDelay || 50,
@@ -338,6 +370,24 @@ class TraceAPI {
     async get(id) {
         return this.client.request(`/api/traces/${id}`);
     }
+    /**
+     * Update an existing trace (e.g. set status, duration, metadata on completion)
+     *
+     * @example
+     * ```typescript
+     * await client.traces.update(42, {
+     *   status: 'success',
+     *   durationMs: 1234,
+     *   metadata: { output: 'done' }
+     * });
+     * ```
+     */
+    async update(id, params) {
+        return this.client.request(`/api/traces/${id}`, {
+            method: 'PATCH',
+            body: JSON.stringify(params),
+        });
+    }
     /**
      * Create a span for a trace
      */

package/dist/index.d.ts

@@ -31,5 +31,6 @@ export { WorkflowTracer, createWorkflowTracer, traceWorkflowStep, traceLangChain
 export type { ClientConfig as AIEvalConfig, Trace as TraceData, Span as SpanData, Evaluation as EvaluationData, LLMJudgeResult as LLMJudgeData, RetryConfig, GenericMetadata as AnnotationData, TracedResponse, TestCase, TestResult, SnapshotData, ExportOptions, ImportOptions, StreamOptions, BatchOptions } from './types';
 export { EvaluationTemplates, type EvaluationTemplateType, type FeatureUsage, type OrganizationLimits } from './types';
 export type { Annotation, CreateAnnotationParams, ListAnnotationsParams, AnnotationTask, CreateAnnotationTaskParams, ListAnnotationTasksParams, AnnotationItem, CreateAnnotationItemParams, ListAnnotationItemsParams, APIKey, APIKeyWithSecret, CreateAPIKeyParams, UpdateAPIKeyParams, ListAPIKeysParams, APIKeyUsage, Webhook, CreateWebhookParams, UpdateWebhookParams, ListWebhooksParams, WebhookDelivery, ListWebhookDeliveriesParams, UsageStats, GetUsageParams, UsageSummary, LLMJudgeConfig, CreateLLMJudgeConfigParams, ListLLMJudgeConfigsParams, ListLLMJudgeResultsParams, LLMJudgeAlignment, GetLLMJudgeAlignmentParams, Organization, } from './types';
+export { parseArgs, runCheck, EXIT, type CheckArgs } from './cli/check';
 import { AIEvalClient } from './client';
 export default AIEvalClient;

package/dist/index.js

@@ -9,7 +9,7 @@
  */
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.decodeCursor = exports.encodeCursor = exports.autoPaginate = exports.createPaginatedIterator = exports.PaginatedIterator = exports.CacheTTL = exports.RequestCache = exports.RateLimiter = exports.batchRead = exports.streamEvaluation = exports.batchProcess = exports.importData = exports.exportData = exports.compareSnapshots = exports.saveSnapshot = exports.compareWithSnapshot = exports.snapshot = exports.TestSuite = exports.createTestSuite = exports.ContextManager = exports.withContext = exports.getContext = exports.createContext = exports.hasValidCodeSyntax = exports.containsAllRequiredFields = exports.followsInstructions = exports.hasNoToxicity = exports.respondedWithinTime = exports.hasFactualAccuracy = exports.containsLanguage = exports.hasReadabilityScore = exports.matchesSchema = exports.hasNoHallucinations = exports.isValidURL = exports.isValidEmail = exports.withinRange = exports.similarTo = exports.hasSentiment = exports.notContainsPII = exports.containsJSON = exports.hasLength = exports.matchesPattern = exports.containsKeywords = exports.expect = exports.NetworkError = exports.ValidationError = exports.AuthenticationError = exports.RateLimitError = exports.EvalAIError = exports.AIEvalClient = void 0;
-exports.EvaluationTemplates = exports.traceAutoGen = exports.traceCrewAI = exports.traceLangChainAgent = exports.traceWorkflowStep = exports.createWorkflowTracer = exports.WorkflowTracer = exports.traceAnthropic = exports.traceOpenAI = exports.Logger = exports.RequestBatcher = void 0;
+exports.EXIT = exports.runCheck = exports.parseArgs = exports.EvaluationTemplates = exports.traceAutoGen = exports.traceCrewAI = exports.traceLangChainAgent = exports.traceWorkflowStep = exports.createWorkflowTracer = exports.WorkflowTracer = exports.traceAnthropic = exports.traceOpenAI = exports.Logger = exports.RequestBatcher = void 0;
 // Main SDK exports
 var client_1 = require("./client");
 Object.defineProperty(exports, "AIEvalClient", { enumerable: true, get: function () { return client_1.AIEvalClient; } });
@@ -106,6 +106,11 @@ Object.defineProperty(exports, "traceAutoGen", { enumerable: true, get: function
 // New exports for v1.1.0
 var types_1 = require("./types");
 Object.defineProperty(exports, "EvaluationTemplates", { enumerable: true, get: function () { return types_1.EvaluationTemplates; } });
+// CLI (programmatic use)
+var check_1 = require("./cli/check");
+Object.defineProperty(exports, "parseArgs", { enumerable: true, get: function () { return check_1.parseArgs; } });
+Object.defineProperty(exports, "runCheck", { enumerable: true, get: function () { return check_1.runCheck; } });
+Object.defineProperty(exports, "EXIT", { enumerable: true, get: function () { return check_1.EXIT; } });
 // Default export for convenience
 const client_2 = require("./client");
 exports.default = client_2.AIEvalClient;

package/dist/types.d.ts

@@ -111,6 +111,14 @@ export interface CreateTraceParams<TMetadata = Record<string, any>> {
     durationMs?: number;
     metadata?: TMetadata;
 }
+/**
+ * Parameters for updating an existing trace
+ */
+export interface UpdateTraceParams<TMetadata = Record<string, any>> {
+    status?: 'pending' | 'success' | 'error';
+    durationMs?: number;
+    metadata?: TMetadata;
+}
 /**
  * Parameters for listing traces
  */

package/dist/workflows.js

@@ -135,13 +135,8 @@ class WorkflowTracer {
         const durationMs = Date.now() - new Date(this.currentWorkflow.startedAt).getTime();
         // Calculate total cost
         const totalCost = this.costs.reduce((sum, cost) => sum + parseFloat(cost.totalCost), 0);
-        // Update the trace with final status
-        // Note: We create a new trace entry with the same ID pattern to update status
-        const traceId = `${this.options.tracePrefix}-complete-${this.currentWorkflow.traceId}`;
-        await this.client.traces.create({
-            name: `Workflow: ${this.currentWorkflow.name}`,
-            traceId,
-            organizationId: this.options.organizationId,
+        // Update the original trace with completion data
+        await this.client.traces.update(this.currentWorkflow.traceId, {
             status: status === 'completed' ? 'success' : 'error',
             durationMs,
             metadata: (0, context_1.mergeWithContext)({