Telemetry: inject dependencies into telemetry app started event

[p-datadog]

What does this PR do?
Injects component tree into telemetry AppStarted event constructor, instead of various methods of the event referencing various global variables.

Motivation:
Make telemetry code easier to read/reason about for the upcoming change to enable telemetry in child processes

Change log entry
None

Additional Notes:
The event does not hold a reference to the component tree - the tree is used only while constructor is running to build the various payloads that will be sent out later.

The unit tests now perform many references to Datadog.send(:components) which looks odd (unit tests reference global state) but this was existing behavior - telemetry was reading global state including in unit tests.

How to test the change?
Existing CI

[github-actions]

Thank you for updating Change log entry section 👏

^{Visited at: 2025-11-19 17:20:24 UTC}

[pr-commenter]

Benchmarks

Benchmark execution time: 2025-11-19 20:03:27

Comparing candidate commit adb655a in PR branch telemetry-inject-deps with baseline commit 276464a in branch master.

Found 1 performance improvements and 0 performance regressions! Performance is the same for 43 metrics, 2 unstable metrics.

scenario:profiling - Allocations ()

• 🟩 throughput [+490864.193op/s; +502640.319op/s] or [+17.312%; +17.727%]

[github-actions]

Typing analysis

Note: Ignored files are excluded from the next sections.

Untyped methods

This PR introduces 5 partially typed methods, and clears 4 partially typed methods. It decreases the percentage of typed methods from 54.67% to 54.64% (-0.03%).

Partially typed methods (+5-4)

❌ Introduced:

sig/datadog/core/telemetry/event/app_started.rbs:10
└── def payload: () -> { products: untyped, configuration: untyped, install_signature: untyped }
sig/datadog/core/telemetry/event/app_started.rbs:17
└── def configuration: (untyped settings, Core::Configuration::AgentSettings agent_settings) -> Array[Hash[Symbol, untyped]]
sig/datadog/core/telemetry/event/app_started.rbs:21
└── def conf_value: (String name, untyped value, Integer seq_id, String origin) -> Hash[Symbol, untyped]
sig/datadog/core/telemetry/event/app_started.rbs:25
└── def install_signature: (untyped settings) -> Hash[Symbol, Object]
sig/datadog/core/telemetry/event/app_started.rbs:27
└── def get_telemetry_origin: (untyped settings, String config_path) -> String

✅ Cleared:

sig/datadog/core/telemetry/event/app_started.rbs:12
└── def payload: () -> { products: untyped, configuration: untyped, install_signature: untyped }
sig/datadog/core/telemetry/event/app_started.rbs:19
└── def configuration: -> Array[Hash[Symbol, untyped]]
sig/datadog/core/telemetry/event/app_started.rbs:23
└── def conf_value: (String name, untyped value, Integer seq_id, String origin) -> Hash[Symbol, untyped]
sig/datadog/core/telemetry/event/app_started.rbs:29
└── def get_telemetry_origin: (untyped config, String config_path) -> String

If you believe a method or an attribute is rightfully untyped or partially typed, you can add # untyped:accept to the end of the line to remove it from the stats.

[datadog-official]

✅ Tests

🎉 All green!

❄️ No new flaky tests detected
🧪 All tests passed

🎯 Code Coverage
• Patch Coverage: 103.77%
• Total Coverage: 98.50%

View detailed report

_{This comment will be updated automatically if new data arrives.
🔗 Commit SHA: 7b333e9 | Docs | Datadog PR Page | Was this helpful? Give us feedback!}

[Strech]

Not sure, maybe it's hard to do a components part in the tests, but I'm not sure that I like the use of Datadog.send(:components) because there is no trace of what's inside.

But this is not a blocker.

[p-datadog]

Agree, all of the references to the component tree are not great. I haven't looked deeply into what it would take to eliminate them -my thinking is it might be tricky because telemetry is pretty much collecting the state of everything else in dd-trace-rb, thus either the test becomes extremely mocky or we have to construct a lot of variables in the tests.