Batch Daily Go Test Parallelizer to process up to 5 files per run

[Copilot]

Daily Go Test Parallelizer was constrained to a single test file per execution, creating unnecessary per-file overhead and limiting throughput. This update shifts the workflow to bounded batching so each daily run can advance multiple files while preserving the same safety posture.

• Batching and throughput

  • Expanded per-run selection from 1 file to a round-robin batch of up to 5 files.
  • Updated state persistence to store the last file in the processed batch so subsequent runs continue correctly.

• Validation cost reduction

  • Replaced per-file validation guidance with deduplicated package-level race runs (go test -race once per unique modified package).
  • Kept a single go test ./... pass after candidate edits to avoid repeated full-suite invocations.

• Safe-output limits and generated workflow

  • Increased create-pull-request.max-patch-files from 1 to 5 to match the new batch size.
  • Regenerated the compiled .lock.yml from the markdown workflow source.

safe-outputs:
  create-pull-request:
    max-patch-files: 5

• Fixes [q] Batch multiple files per run in Daily Go Test Parallelizer #49955

[pelikhan]

@copilot optimize with inline agents to preserve small context, process up to 25 files.

[Copilot]

@copilot optimize with inline agents to preserve small context, process up to 25 files.

Updated in d41a816. The workflow now processes up to 25 files per run and adds inline parallel-safety-checker agent guidance to keep per-file analysis context small, with regenerated .lock.yml.

[Copilot]

Pull request overview

Updates the daily Go test parallelizer to batch files, delegate safety checks, and deduplicate validation.

Changes:

• Expands round-robin batching and safe-output limits.
• Adds inline safety-analysis agents.
• Updates token-optimization skill references.

Show a summary per file

File	Description
.github/workflows/daily-go-test-parallelizer.md	Implements batching, sub-agents, validation, and state persistence.
.github/workflows/daily-go-test-parallelizer.lock.yml	Regenerates the compiled workflow.
.github/skills/agentic-workflows/SKILL.md	Adds an observability reference.

Review details

Tip

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

Suppressed comments (2)

.github/workflows/daily-go-test-parallelizer.md:50

• The PR title and description specify a batch of up to five files, but the workflow tells the agent to analyze twenty-five. This materially increases cost, timeout risk, and PR scope beyond the requested bound. Change this to five and regenerate the lock file.

Analyze up to twenty-five Go test files per run and add `t.Parallel()` only where parallel execution is demonstrably safe.

.github/workflows/daily-go-test-parallelizer.md:58

• These selection limits also implement a 25-file batch instead of the five-file batch promised by the PR. Cap both values at 5 so round-robin advancement and the safe-output limit describe the same intended workload.

3. Select up to 25 consecutive paths starting after `last_file`, wrapping to the first path. If the cache is absent, malformed, or names a removed file, start from the first path.
4. Analyze and modify only this selected batch (maximum 25 files).

• Files reviewed: 3/3 changed files
• Comments generated: 3
• Review effort level: Balanced

[github-actions]

🎉 This pull request is included in a new release.

Release: v0.84.4