✨ Add datadog-sdk-event-inspection skill for Claude Code

[mormubis]

Motivation

Adds a Claude Code skill for inspecting Browser SDK event payloads in the sandbox. This provides a reusable, documented workflow for verifying emitted events without relying on network request interception.

Changes

• Adds .claude/skills/datadog-sdk-event-inspection/SKILL.md — a skill that guides Claude Code to use the window.__ddBrowserSdkExtensionCallback hook for event inspection, with two injection paths (console post-load vs CDP pre-load for initial view events).

Test instructions

Load the skill in a Claude Code session and verify it correctly guides event inspection in the sandbox (yarn dev → http://localhost:8080).

Checklist

• Tested locally
• Tested on staging
• Added unit tests for this change.
• Added e2e/integration tests for this change.
• Updated documentation and/or relevant AGENTS.md file

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: ebd1ce9c1e

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[cit-pr-commenter-54b7da]

Bundles Sizes Evolution

📦 Bundle Name	Base Size	Local Size	𝚫	𝚫%	Status
Rum	174.43 KiB	174.43 KiB	0 B	0.00%	✅
Rum Profiler	6.16 KiB	6.16 KiB	0 B	0.00%	✅
Rum Recorder	25.24 KiB	25.24 KiB	0 B	0.00%	✅
Logs	56.84 KiB	56.84 KiB	0 B	0.00%	✅
Flagging	944 B	944 B	0 B	0.00%	✅
Rum Slim	130.11 KiB	130.11 KiB	0 B	0.00%	✅
Worker	23.63 KiB	23.63 KiB	0 B	0.00%	✅

🚀 CPU Performance

Action Name	Base CPU Time (ms)	Local CPU Time (ms)	𝚫%
RUM - add global context	0.0094	0.0042	-55.32%
RUM - add action	0.0489	0.0144	-70.55%
RUM - add error	0.0379	0.0137	-63.85%
RUM - add timing	0.0061	0.0029	-52.46%
RUM - start view	0.0217	0.0117	-46.08%
RUM - start/stop session replay recording	0.0012	0.0007	-41.67%
Logs - log message	0.0283	0.0182	-35.69%

🧠 Memory Performance

Action Name	Base Memory Consumption	Local Memory Consumption	𝚫
RUM - add global context	25.95 KiB	26.60 KiB	+664 B
RUM - add action	50.84 KiB	51.59 KiB	+770 B
RUM - add timing	26.68 KiB	26.44 KiB	-245 B
RUM - add error	56.47 KiB	54.13 KiB	-2.34 KiB
RUM - start/stop session replay recording	25.80 KiB	25.94 KiB	+137 B
RUM - start view	451.32 KiB	450.04 KiB	-1.28 KiB
Logs - log message	44.45 KiB	44.87 KiB	+435 B

🔗 RealWorld

[datadog-datadog-prod-us1]

✅ Tests

🎉 All green!

❄️ No new flaky tests detected
🧪 All tests passed

🎯 Code Coverage (details)
• Patch Coverage: 100.00%
• Overall Coverage: 76.76% (+0.01%)

_{This comment will be updated automatically if new data arrives.
🔗 Commit SHA: 2cc1269 | Docs | Datadog PR Page | Was this helpful? React with 👍/👎 or give us feedback!}

[thomas-lebeau]

❓ question: can we use mermaid diagram instead? they are commonly rendered in markdown previews

```mermaid
flowchart TD
    Q["Need initial view event<br/>or events from init()?"]
    Q -- "yes" --> B["Path B: CDP pre-load injection"]
    Q -- "no" --> A["Path A: Console injection"]

[mormubis]

Already applied — the mermaid flowchart is in the skill file (commit 279517d).

[thomas-lebeau]

💭 thought: ‏I found the sandbox useful for reproducing customer escalations. However, if unit test is not enough, then maybe what we want is an e2e test, which is doing the same thing but stays in the repo to prevent regression.

[mormubis]

Good point. Added a note to AGENTS.md: 'this skill is a debugging and spot-check tool, not a regression gate. If you need to prevent regressions, write an E2E test instead (see test/e2e/).' (commit 3ec3534)

[thomas-lebeau]

🔨 warning: ‏It's missing mention of Jasmine and Karma. Without that Claude used to be confused in test APIs

[mormubis]

Already there — AGENTS.md has - **Test framework**: Jasmine + Karma in the Unit Tests section.