Add AWF/AWMG runtime versions to OTLP resource attributes

[Copilot]

OTLP spans already carried the gh-aw CLI service.version, but not the AWF and MCP gateway runtime versions that actually executed the workflow. That left incident correlation blind to runtime version bumps even though both values were already written to /tmp/gh-aw/aw_info.json.

• What changed

  • Extended shared OTLP resource attribute construction to emit:
    • gh-aw.awf.version
    • gh-aw.awmg.version
  • Read both values from aw_info.json with env-var fallbacks for cases where runtime metadata is not available from disk.

• Where it now applies

  • Job setup spans in send_otlp_span.cjs
  • Job conclusion spans in send_otlp_span.cjs
  • Generic tool spans in otlp.cjs

• Why this matters

  • Makes runtime incident analysis distinguish:
    • gh-aw CLI version (service.version)
    • AWF runtime version (gh-aw.awf.version)
    • MCP gateway version (gh-aw.awmg.version)
  • Enables direct filtering and correlation when workflow behavior changes after AWF/AWMG bumps.

• Coverage added

  • Added focused tests to verify the new resource attributes are propagated from aw_info.json for setup spans, conclusion spans, and generic OTLP tool spans.

constresourceAttributes=buildGitHubActionsResourceAttributes({
repository,
runId,
awfVersion,
awmgVersion,
staged,
runAttempt,});

[pelikhan]

@copilot recompile and ensure the version envs are set in the compiled lock file

[Copilot]

@copilot recompile and ensure the version envs are set in the compiled lock file

I recompiled smoke-otel-backends locally and rechecked the generated lock; it stays clean because the compiled workflow already emits GH_AW_INFO_AWF_VERSION / GH_AW_INFO_AWMG_VERSION in the lock file. The OTLP runtime change remains in c383626.

[Copilot]

Pull request overview

This PR extends OTLP resource attribute generation to include the AWF runtime and MCP gateway runtime versions (sourced from /tmp/gh-aw/aw_info.json with env-var fallbacks) so traces can be correlated against runtime version changes.

Changes:

• Added gh-aw.awf.version and gh-aw.awmg.version to shared GitHub Actions OTLP resource attributes.
• Plumbed awf_version / awmg_version from aw_info.json (with GH_AW_INFO_AWF_VERSION / GH_AW_INFO_AWMG_VERSION fallbacks) into setup, conclusion, and generic tool spans.
• Added targeted tests to validate the new resource attributes across span types.

Show a summary per file

File	Description
pkg/workflow/gh_cli_permissions.go	Formatting/alignment cleanup in JSON struct definitions.
actions/setup/js/send_otlp_span.cjs	Emits AWF/AWMG runtime versions as OTLP resource attributes and wires in aw_info/env fallbacks for setup & conclusion spans.
actions/setup/js/send_otlp_span.test.cjs	Adds assertions that setup and conclusion span resources include gh-aw.awf.version / gh-aw.awmg.version.
actions/setup/js/otlp.cjs	Passes AWF/AWMG runtime versions through to shared resource-attribute builder for generic tool spans.
actions/setup/js/otlp.test.cjs	Adds assertion that logSpan forwards AWF/AWMG versions from aw_info.json into resource attribute construction.

Copilot's findings

Tip

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

Comments suppressed due to low confidence (1)

actions/setup/js/send_otlp_span.test.cjs:4805

• The sendJobConclusionSpan tests save/restore a fixed envKeys list, but it doesn't clear the new GH_AW_INFO_AWF_VERSION / GH_AW_INFO_AWMG_VERSION env-var fallbacks. If either is set in the ambient environment, conclusion-span tests (including this new one) can become flaky. Include these vars in envKeys (or delete/restore them in hooks) to keep tests deterministic.

 it("includes gh-aw.awf.version and gh-aw.awmg.version on conclusion span resources from aw_info.json", async () => {
const mockFetch = vi.fn().mockResolvedValue({ ok: true, status: 200, statusText: "OK" });
vi.stubGlobal("fetch", mockFetch);
process.env.GH_AW_OTLP_ENDPOINTS = JSON.stringify([{ url: "https://traces.example.com" }]);

• Files reviewed: 4/5 changed files
• Comments generated: 2