Improve JS step-summary agent trace readability with ordered event markers

[Copilot]

The step summary currently renders agent logs as a flat stream, which makes trace progression hard to scan during workflow triage. This update improves JavaScript log-trace rendering by introducing explicit event ordering inspired by agent-trace timeline ergonomics.

• Problem focus

  • Improve readability of agent log traces in GitHub step summaries without changing execution semantics.

• Rendering changes

  • Ordered trace markers: assistant reasoning blocks and tool execution lines now carry monotonic [n] markers.
  • Stable output shape: existing compact result previews (├/└ lines, truncation rules, tool status icons) are preserved; only scanability is improved.

• Test updates

  • Extended summary-rendering assertions to verify numbered trace output for both success and error tool paths.

if (i === 0) {
  state.traceEventCount += 1;
}
const prefix = i === 0 ? `[${state.traceEventCount}] ◆ ` : "  ";

// ...

state.traceEventCount += 1;
appendConversationLine(lines, `[${state.traceEventCount}] ${statusIcon} ${displayName}`, state);

[pelikhan]

@copilot render a preview for https://github.com/github/gh-aw/actions/runs/25953787841/job/76296776802

[pelikhan]

@copilot show time deltas like in npmjs debug package

[Copilot]

@copilot render a preview for https://github.com/github/gh-aw/actions/runs/25953787841/job/76296776802

Preview from that job’s rendered agent summary:

=== Codex Execution Summary ===
Model: gpt-5.4-mini

Conversation:

Statistics:

In this run, no assistant/tool trace lines were emitted into the conversation section.

[Copilot]

Pull request overview

Adds monotonic [n] event markers to assistant text blocks and tool execution lines in the JS log parser's step-summary output to improve scanability of agent traces. Existing output structure (icons, preview lines, truncation) is preserved.

Changes:

• Introduce traceEventCount in the summary rendering state and increment it for each agent text block and tool execution line.
• Prepend [n] markers to the first line of agent text and to tool execution lines.
• Extend tests to assert the new numbered markers for both success and error tool paths.

Show a summary per file

File	Description
actions/setup/js/log_parser_format.cjs	Adds traceEventCount state and prefixes agent text and tool execution lines with [n] markers.
actions/setup/js/log_parser_shared.test.cjs	Adds assertions for the new [n] markers on agent text and tool lines in success/error paths.

Copilot's findings

Tip

Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

• Files reviewed: 2/2 changed files
• Comments generated: 0