Merge remote-tracking branch 'upstream/main' into more-pinning

# Conflicts:
#	instrumentation/kubernetes-client-7.0/javaagent/build.gradle.kts
#	instrumentation/pekko/pekko-http-1.0/javaagent/build.gradle.kts

Author: trask

.fossa.yml

@@ -367,9 +367,6 @@ targets:
     - type: gradle
       path: ./
       target: ':instrumentation:aws-lambda:aws-lambda-events-2.2:javaagent'
-    - type: gradle
-      path: ./
-      target: ':instrumentation:aws-lambda:aws-lambda-events-2.2:library'
     - type: gradle
       path: ./
       target: ':instrumentation:aws-lambda:aws-lambda-events-3.11:library'
@@ -570,7 +567,7 @@ targets:
       target: ':instrumentation:jaxws:jaxws-2.0-metro-2.2:javaagent'
     - type: gradle
       path: ./
-      target: ':instrumentation:jaxws:jaxws-common:javaagent'
+      target: ':instrumentation:jaxws:jaxws-common-2.0:javaagent'
     - type: gradle
       path: ./
       target: ':instrumentation:jaxws:jaxws-jws-api-1.1:javaagent'
@@ -1053,7 +1050,7 @@ targets:
       target: ':instrumentation:xxl-job:xxl-job-2.3.0:javaagent'
     - type: gradle
       path: ./
-      target: ':instrumentation:xxl-job:xxl-job-common:javaagent'
+      target: ':instrumentation:xxl-job:xxl-job-common-1.9.2:javaagent'
     - type: gradle
       path: ./
       target: ':instrumentation:zio:zio-2.0:javaagent'

.github/agents/knowledge/gradle-conventions.md

@@ -266,20 +266,27 @@ These system properties support the metadata collection pipeline. They are not r
 test correctness and are being added as a separate migration — **do not add them during
 review**. Only verify correctness when they are already present.
 
+Do not add `collectMetadata` or `metadataConfig` to `javaagent-unit-tests` projects. These are
+unit tests, and metadata collection should not run there.
+
 | Property | Type | Value |
 | --- | --- | --- |
 | `collectMetadata` | System property | Pass-through of `otelProps.collectMetadata`; defaults to `false` |
 | `metadataConfig` | System property | A single `key=value` string describing the non-default configuration active during this test run |
 
 When already present, verify:
 
+- `metadataConfig` is only used in files that also configure `collectMetadata`. A lone
+  `metadataConfig` does not enable collection and should be removed, not added as a partial
+  metadata migration.
 - `collectMetadata` is in `tasks.test` for single-test-task modules, or in
   `withType<Test>().configureEach` for modules that explicitly register additional `Test`
   tasks via `by registering(Test::class)` (`latestDepTest` does not count) — never on
   individual tasks. Do not use
   `withType<Test>().configureEach { ... }` in single-test-task modules.
-- `metadataConfig` is on each non-default task. It may also appear on the default `test`
-  task when that task itself runs with non-default `jvmArgs` (e.g., an experimental flag
-  enabled module-wide via `withType<Test>().configureEach { jvmArgs(...) }`); in that case
-  the `metadataConfig` value should describe those non-default jvmArgs.
+- `metadataConfig` is on each non-default task that participates in metadata collection. It may
+  also appear on the default `test` task when that task participates in metadata collection and
+  itself runs with non-default `jvmArgs` (e.g., an experimental flag enabled module-wide via
+  `withType<Test>().configureEach { jvmArgs(...) }`); in that case the `metadataConfig` value
+  should describe those non-default jvmArgs.
 - The `metadataConfig` value matches at least one of the jvmArgs configured in the task

.github/agents/knowledge/javaagent-advice-patterns.md

@@ -42,7 +42,9 @@ the IDE's perspective, and per-method placement is inconsistent with the rest of
 ## Advice Methods Must Be Static
 
 All `@Advice.OnMethodEnter` and `@Advice.OnMethodExit` methods **must be `static`**. ByteBuddy
-inlines advice code directly into the instrumented method — there is no advice object instance.
+inlines advice code directly into the instrumented method by default — there is no advice object
+instance. Do not add or remove `inline = false` as cleanup; it is a semantic signal used by
+muzzle/indy tooling.
 
 ```java
 // ✅ Correct

.github/agents/knowledge/metadata-yaml-format.md

@@ -87,6 +87,12 @@ Add `examples` only for module-specific configs with non-obvious format (lists,
 ### 1. Validate experimental markers match
 
 - `/development` in declarative_name ↔ `experimental` in flat name (MUST match both ways)
+- Do not rename an existing, published `declarative_name` solely to make it match the mechanical
+  conversion rule. Flat property lookup normalizes `-` to `.`, so a legacy YAML key such as
+  `java.aws_sdk.use_propagator_for_messaging/development` can still resolve to
+  `otel.instrumentation.aws-sdk.experimental-use-propagator-for-messaging` without a special
+  mapping. Let the automated validation decide whether a `SPECIAL_MAPPINGS` entry is actually
+  needed.
 - WRONG: `otel.instrumentation.servlet.capture-request-parameters` + `java.servlet.capture_request_parameters/development`
 - RIGHT: `otel.instrumentation.servlet.experimental.capture-request-parameters` + `java.servlet.capture_request_parameters/development`
 
@@ -152,4 +158,6 @@ FAIL in ../instrumentation/liberty/liberty-20.0/metadata.yaml:
 ## Edge Cases
 
 - Properties without `otel.instrumentation.` prefix → check SPECIAL_MAPPINGS
-- Already has declarative_name → skip conversion
+- Already has declarative_name → skip conversion unless the automated validation fails; preserve
+  existing user-facing YAML keys and add a `SPECIAL_MAPPINGS` bridge entry only when normalized flat
+  property lookup cannot resolve the existing name

.github/agents/knowledge/testing-general-patterns.md

@@ -30,6 +30,11 @@
   non-checked wait path exists (for example, when a timeout is required via
   `get(timeout, unit)`), leave the test's `throws` clause as-is — including `throws Exception`
   — rather than inventing a new helper just to narrow it.
+- Do **not** replace a direct `CountDownLatch.await(timeout, unit)` assertion with Awaitility
+  polling of `getCount()` solely to avoid `InterruptedException` and narrow the test method's
+  `throws` clause. Prefer the conventional
+  `assertThat(latch.await(timeout, unit)).isTrue()` form for latch-based callback waits, even
+  when that means leaving `throws Exception` on the test method.
 - Do **not** wrap a checked exception inside a lambda body (for example, catching
   `IOException` and rethrowing `UncheckedIOException`) solely to narrow a test method's
   `throws` clause. That noisy try/catch inside the lambda is worse than leaving
@@ -107,9 +112,12 @@
   valid until this method returns; for example, if the assertion expects a `used` connection point,
   keep the borrowed connection open until after the assertion.
 - After removing a metric-producing source or unregistering an observable callback, call
-  `testing().clearData()` and then use `waitAndAssertMetrics(..., AbstractIterableAssert::isEmpty)`
-  for absence checks. Do not add an exporter-interval sleep before or after `clearData()` solely to
-  wait for metrics; the test runners force-flush metrics when reading them.
+  `testing().clearData()` and then use Awaitility around the simplest assertion that captures the
+  intent. For example, when the test expects no metrics from an instrumentation scope, prefer
+  `await().untilAsserted(() -> assertThat(testing().metrics()).filteredOn(...).isEmpty())` over
+  expanding the check into one `waitAndAssertMetrics(..., AbstractIterableAssert::isEmpty)` call per
+  possible metric name. Do not add an exporter-interval sleep before or after `clearData()` solely
+  to wait for metrics; the test runners force-flush metrics when reading them.
 
 ## Attribute Assertion `satisfies()` Lambda Parameters
 

.github/agents/knowledge/testing-semconv-stability.md

@@ -3,7 +3,7 @@
 ## Quick Reference
 
 - Use when: reviewing semconv dual-mode assertions or `testStableSemconv` / `testBothSemconv` tasks
-- Review focus: mode-specific assertions, `maybeStable()` usage boundaries, class-level deprecation suppression
+- Review focus: mode-specific assertions, `maybeStable()` usage boundaries
 
 ## Background: The Three Modes
 
@@ -109,7 +109,3 @@ if (emitOldCodeSemconv()) {
   assertThat(attributes).containsEntry(CODE_NAMESPACE, "MyClass");
 }
 ```
-
-## Key Rules
-
-- Add `@SuppressWarnings("deprecation")` at class level when tests use old Semconv constants.

.github/agents/module-cleanup.agent.md

@@ -113,9 +113,9 @@ Auto-fix boundaries:
     whose directory name contains `-common`, or whose Gradle path ends with `:testing`,
     `:library`, or `:bootstrap`, **first search all sibling modules** under the same
     instrumentation parent for callers of the type or member before applying the fix.
-  - normalization of existing `@SuppressWarnings` syntax or placement, but preserve any
-    accurate explanatory comment attached to the suppression instead of deleting it as
-    style noise
+  - normalization of existing `@SuppressWarnings` syntax or placement, but only when it preserves
+    or narrows the suppression to the minimal required scope; preserve any accurate explanatory
+    comment attached to the suppression instead of deleting it as style noise
   - obvious assertion API migrations (e.g., AssertJ preference) and idiomatic
     simplifications listed in `testing-general-patterns.md` § AssertJ Idiomatic
     Simplifications (e.g., `assertThat(list.size()).isEqualTo(N)` →

.github/agents/pr-review.agent.md

@@ -119,9 +119,9 @@
     "commons-httpclient:commons-httpclient#+": "3.1",
     "commons-httpclient:commons-httpclient#3.+": "3.1",
     "dev.failsafe:failsafe#+": "3.3.2",
-    "dev.zio:zio-http_2.12#+": "3.11.0",
-    "dev.zio:zio-http_2.13#+": "3.11.0",
-    "dev.zio:zio-http_3#+": "3.11.0",
+    "dev.zio:zio-http_2.12#+": "3.11.1",
+    "dev.zio:zio-http_2.13#+": "3.11.1",
+    "dev.zio:zio-http_3#+": "3.11.1",
     "dev.zio:zio_2.12#+": "2.1.25",
     "dev.zio:zio_2.13#+": "2.1.25",
     "dev.zio:zio_3#+": "2.1.25",

.github/config/latest-dep-versions.json

@@ -7,7 +7,7 @@ max_retries = 6
 max_concurrency = 4
 
 # Check link anchors
-include_fragments = true
+include_fragments = "full"
 
 remap = [
   "https://docs.oracle.com/(.*)#.*$ https://docs.oracle.com/$1"