fix remaining issues

Author: rochdev

.agents/skills/llmobs-integration/SKILL.md

@@ -93,16 +93,16 @@ See [references/plugin-architecture.md](references/plugin-architecture.md) for c
 Answer these questions by reading the code:
 
 1. **Does the package make direct HTTP calls to LLM provider endpoints?**
-   - YES → Go to question 2
-   - NO → Go to question 3
+    - YES → Go to question 2
+    - NO → Go to question 3
 
 2. **Does it support multiple LLM providers via configuration?**
-   - YES → **`LlmObsCategory.MULTI_PROVIDER`**
-   - NO → **`LlmObsCategory.LLM_CLIENT`**
+    - YES → **`LlmObsCategory.MULTI_PROVIDER`**
+    - NO → **`LlmObsCategory.LLM_CLIENT`**
 
 3. **Does it implement workflow/graph orchestration with state management?**
-   - YES → **`LlmObsCategory.ORCHESTRATION`**
-   - NO → **`LlmObsCategory.INFRASTRUCTURE`**
+    - YES → **`LlmObsCategory.ORCHESTRATION`**
+    - NO → **`LlmObsCategory.INFRASTRUCTURE`**
 
 See [references/category-detection.md](references/category-detection.md) for detailed heuristics and examples.
 
@@ -138,31 +138,31 @@ See [references/message-extraction.md](references/message-extraction.md) for pro
 ## Implementation Steps
 
 1. **Detect package category** (REQUIRED FIRST STEP)
-   - Follow decision tree above
-   - Output: category, confidence, reasoning
+    - Follow decision tree above
+    - Output: category, confidence, reasoning
 
 2. **Create plugin file**
-   - Location: `packages/dd-trace/src/llmobs/plugins/{integration}/index.js`
-   - Extend: `LLMObsPlugin` base class
-   - Implement: Required methods per plugin architecture
+    - Location: `packages/dd-trace/src/llmobs/plugins/{integration}/index.js`
+    - Extend: `LLMObsPlugin` base class
+    - Implement: Required methods per plugin architecture
 
 3. **Implement `getLLMObsSpanRegisterOptions(ctx)`**
-   - Extract model provider and name from context
-   - Determine span kind (usually `'llm'`)
-   - Return registration options object
+    - Extract model provider and name from context
+    - Determine span kind (usually `'llm'`)
+    - Return registration options object
 
 4. **Implement `setLLMObsTags(ctx)`**
-   - Extract input messages from `ctx.arguments`
-   - Extract output messages from `ctx.result`
-   - Extract token metrics (input_tokens, output_tokens, total_tokens)
-   - Extract metadata (temperature, max_tokens, etc.)
-   - Tag span using `this._tagger` methods
+    - Extract input messages from `ctx.arguments`
+    - Extract output messages from `ctx.result`
+    - Extract token metrics (input_tokens, output_tokens, total_tokens)
+    - Extract metadata (temperature, max_tokens, etc.)
+    - Tag span using `this._tagger` methods
 
 5. **Handle edge cases**
-   - Streaming responses (if applicable)
-   - Error cases (empty output messages)
-   - Non-standard message formats
-   - Missing metadata
+    - Streaming responses (if applicable)
+    - Error cases (empty output messages)
+    - Non-standard message formats
+    - Missing metadata
 
 See [references/plugin-architecture.md](references/plugin-architecture.md) for step-by-step implementation guide.
 

.agents/skills/llmobs-integration/references/category-detection.md

@@ -87,16 +87,16 @@ Follow this tree to determine category:
 
 ```
 1. Does the package make direct HTTP calls to LLM provider endpoints?
-   ├─ YES → Go to question 2
-   └─ NO  → Go to question 3
+    ├─ YES → Go to question 2
+    └─ NO  → Go to question 3
 
 2. Does it support multiple LLM providers via configuration?
-   ├─ YES → LlmObsCategory.MULTI_PROVIDER
-   └─ NO  → LlmObsCategory.LLM_CLIENT
+    ├─ YES → LlmObsCategory.MULTI_PROVIDER
+    └─ NO  → LlmObsCategory.LLM_CLIENT
 
 3. Does it implement workflow/graph orchestration with state management?
-   ├─ YES → LlmObsCategory.ORCHESTRATION
-   └─ NO  → LlmObsCategory.INFRASTRUCTURE
+    ├─ YES → LlmObsCategory.ORCHESTRATION
+    └─ NO  → LlmObsCategory.INFRASTRUCTURE
 ```
 
 ## Detection Process

.agents/skills/llmobs-testing/references/assertion-helpers.md

@@ -99,34 +99,34 @@ assertLlmObsSpanEvent(events[0], {
 ## Best Practices
 
 1. **Use MOCK_* for non-deterministic values:**
-   - Output text: `MOCK_STRING` (real responses vary)
-   - Token counts: `MOCK_NOT_NULLISH` (counts vary but should exist)
-   - Error objects: `MOCK_OBJECT` (error details vary)
+    - Output text: `MOCK_STRING` (real responses vary)
+    - Token counts: `MOCK_NOT_NULLISH` (counts vary but should exist)
+    - Error objects: `MOCK_OBJECT` (error details vary)
 
 2. **Use exact values for inputs:**
-   - Input messages: You control these in tests
-   - Model parameters: You set these (temperature, max_tokens)
-   - Model name: You specify this
+    - Input messages: You control these in tests
+    - Model parameters: You set these (temperature, max_tokens)
+    - Model name: You specify this
 
 3. **Always validate core fields:**
-   - `spanKind` (required for every span)
-   - `name` (operation identifier)
-   - `modelName` and `modelProvider` (for LLM spans)
+    - `spanKind` (required for every span)
+    - `name` (operation identifier)
+    - `modelName` and `modelProvider` (for LLM spans)
 
 4. **Validate message format:**
-   - Ensure `{content: string, role: string}` structure
-   - Check role values: `'user'`, `'assistant'`, `'system'`, `'tool'`
+    - Ensure `{content: string, role: string}` structure
+    - Check role values: `'user'`, `'assistant'`, `'system'`, `'tool'`
 
 5. **Test error paths:**
-   - Verify empty `outputMessages: [{content: '', role: ''}]` on errors
-   - Assert `error` field exists with `MOCK_OBJECT`
+    - Verify empty `outputMessages: [{content: '', role: ''}]` on errors
+    - Assert `error` field exists with `MOCK_OBJECT`
 
 6. **Match span kind to operation:**
-   - Chat/completions → `spanKind: 'llm'`
-   - Workflow execution → `spanKind: 'workflow'`
-   - Agent runs → `spanKind: 'agent'`
-   - Tool calls → `spanKind: 'tool'`
-   - Embeddings → `spanKind: 'embedding'`
+    - Chat/completions → `spanKind: 'llm'`
+    - Workflow execution → `spanKind: 'workflow'`
+    - Agent runs → `spanKind: 'agent'`
+    - Tool calls → `spanKind: 'tool'`
+    - Embeddings → `spanKind: 'embedding'`
 
 ## Reference Test Implementation
 

.editorconfig-checker.json

@@ -7,6 +7,7 @@
   "Exclude": [
     "LICENSE",
     "LICENSE.*",
+    "packages/dd-trace/test/llmobs/cassettes",
     "vendor"
   ],
   "AllowedContentTypes": [],

eslint-rules/eslint-env-aliases.test.mjs

@@ -21,15 +21,15 @@ ruleTester.run('eslint-env-aliases', rule, {
       output: "const env = 'DD_PROFILING_CPU_ENABLED'",
       errors: [{
         message: "Use canonical environment variable name 'DD_PROFILING_CPU_ENABLED' " +
-                 "instead of alias 'DD_PROFILING_EXPERIMENTAL_CPU_ENABLED'",
+          "instead of alias 'DD_PROFILING_EXPERIMENTAL_CPU_ENABLED'",
       }],
     },
     {
       code: 'const env = `DD_TRACE_EXPERIMENTAL_RUNTIME_ID_ENABLED`',
       output: "const env = 'DD_RUNTIME_METRICS_RUNTIME_ID_ENABLED'",
       errors: [{
         message: "Use canonical environment variable name 'DD_RUNTIME_METRICS_RUNTIME_ID_ENABLED' " +
-                 "instead of alias 'DD_TRACE_EXPERIMENTAL_RUNTIME_ID_ENABLED'",
+          "instead of alias 'DD_TRACE_EXPERIMENTAL_RUNTIME_ID_ENABLED'",
       }],
     },
   ],

integration-tests/coverage/README.md

@@ -20,27 +20,27 @@ All sources live in `integration-tests/coverage/`.
 
 ```md
 npm run test:integration:mocha:coverage
-   │
-   ▼
+  │
+  ▼
 integration-tests/coverage/run-suite.js
-   │  (spawns mocha with --require register.js)
-   ▼
+  │  (spawns mocha with --require register.js)
+  ▼
 register.js  ──▶ patch-child-process.js  (monkey-patches child_process)
-   │
-   ▼  createSandbox → packTarballWithLock
+  │
+  ▼  createSandbox → packTarballWithLock
 pack-instrumented-tarball.js
-   │  (bun pm pack → extract → istanbul instrument → sentinel → tar -czf)
-   ▼  each spawn/fork/exec of Node
+  │  (bun pm pack → extract → istanbul instrument → sentinel → tar -czf)
+  ▼  each spawn/fork/exec of Node
 child-bootstrap.js  ──▶ require('nyc').wrap() [hookRequire: false]
-   │  (child runs, writes raw coverage JSON to per-sandbox temp dir)
-   ▼
+  │  (child runs, writes raw coverage JSON to per-sandbox temp dir)
+  ▼
 useSandbox teardown
-   │
-   ▼
+  │
+  ▼
 finalize-sandbox.js  (reads sandbox JSON, rebases paths to repo root,
-                     writes coverage into the collector)
-   │
-   ▼
+                    writes coverage into the collector)
+  │
+  ▼
 merge-lcov.js  (called by run-suite.js after mocha exits; merges every
                 sandbox into coverage/node-<version>-<script>/lcov.info)
 ```

integration-tests/jest/jest.spec.js

@@ -7479,11 +7479,11 @@ describe(`jest@${JEST_VERSION} commonJS`, () => {
         fs.writeFileSync(
           path.join(cwd, 'ci-visibility/test-management/test-worker-restart-z-impacted.js'),
           `const assert = require('assert')
-           describe('worker restart impacted tests', () => {
-             it('can pass normally', () => {
-               assert.strictEqual(1 + 2, 3)
-             })
-           })`
+          describe('worker restart impacted tests', () => {
+            it('can pass normally', () => {
+              assert.strictEqual(1 + 2, 3)
+            })
+          })`
         )
         execSync('git add ci-visibility/test-management/test-worker-restart-z-impacted.js', { cwd, stdio: 'ignore' })
         execSync('git commit --amend --no-edit', { cwd, stdio: 'ignore' })

integration-tests/playwright/playwright-impacted-tests.spec.js

@@ -105,24 +105,24 @@ versions.forEach((version) => {
           path.join(cwd, 'ci-visibility/playwright-tests-impacted-tests/impacted-test.js'),
           `const { test, expect } = require('@playwright/test')
 
-           test.beforeEach(async ({ page }) => {
-             await page.goto(process.env.PW_BASE_URL)
-           })
-
-           test.describe('impacted test', () => {
-             test('should be impacted', async ({ page }) => {
-               await expect(page.locator('.hello-world')).toHaveText([
-                 'Hello Worldd'
-               ])
-             })
-           })
-           test.describe('impacted test 2', () => {
-             test('should be impacted 2', async ({ page }) => {
-               await expect(page.locator('.hello-world')).toHaveText([
-                 'Hello World'
-               ])
-             })
-           })`
+          test.beforeEach(async ({ page }) => {
+            await page.goto(process.env.PW_BASE_URL)
+          })
+
+          test.describe('impacted test', () => {
+            test('should be impacted', async ({ page }) => {
+              await expect(page.locator('.hello-world')).toHaveText([
+                'Hello Worldd'
+              ])
+            })
+          })
+          test.describe('impacted test 2', () => {
+            test('should be impacted 2', async ({ page }) => {
+              await expect(page.locator('.hello-world')).toHaveText([
+                'Hello World'
+              ])
+            })
+          })`
         )
         execSync('git add ci-visibility/playwright-tests-impacted-tests/impacted-test.js', { cwd, stdio: 'ignore' })
         execSync('git commit -m "modify impacted-test.js" --no-verify', { cwd, stdio: 'ignore' })

packages/datadog-instrumentations/src/mocha/main.js

@@ -692,9 +692,9 @@ addHook({
   shimmer.wrap(BufferedWorkerPool.prototype, 'run', run => async function (testSuiteAbsolutePath, workerArgs) {
     if (!testFinishCh.hasSubscribers ||
         (!config.isKnownTestsEnabled &&
-         !config.isTestManagementTestsEnabled &&
-         !config.isImpactedTestsEnabled &&
-         !config.isFlakyTestRetriesEnabled)) {
+          !config.isTestManagementTestsEnabled &&
+          !config.isImpactedTestsEnabled &&
+          !config.isFlakyTestRetriesEnabled)) {
       return run.apply(this, arguments)
     }
 

packages/dd-trace/src/opentracing/propagation/text_map.js

@@ -818,7 +818,7 @@ class TextMapPropagator {
           (!tracestateSamplingPriority || tracestateSamplingPriority >= 0)) {
         samplingPriority = AUTO_REJECT
       } else if (traceparentSampled === 1 &&
-                 (!tracestateSamplingPriority || tracestateSamplingPriority < 0)) {
+        (!tracestateSamplingPriority || tracestateSamplingPriority < 0)) {
         samplingPriority = AUTO_KEEP
       }
     }