Description
engine: claude (or nested id: claude) appears 58 times across .github/workflows/*.md versus 115 for Copilot (~1.98x), and Claude-specific smoke-test variants number only 9 (smoke-claude*.md, smoke-github-claude.md, smoke-claude-on-copilot.md) versus 15 Copilot variants (arm64, SDK, sub-agents, AOAI apikey/Entra auth modes, small-model, etc). A Claude-only user has roughly half the reference material to copy from when building a new workflow, and several auth/runtime modes Copilot covers (sub-agents, ARM64, alternate auth) have no Claude equivalent smoke test.
Expected Impact
Closes the reference-material gap for Claude users and gives CI equivalent coverage of Claude-specific runtime modes (ARM64, SDK-style invocation, sub-agents, alternate auth) that currently only Copilot smoke tests exercise.
Suggested Agent
GitHub Copilot coding agent or a workflow-authoring agent — add 4-6 new smoke-claude-*.md workflows mirroring the Copilot variants (e.g. smoke-claude-arm64.md, smoke-claude-sdk.md, smoke-claude-sub-agents.md, an alternate-auth-mode variant), following the existing smoke-claude.md / smoke-copilot-*.md structure, then run make recompile.
Estimated Effort
Medium (1-4 hours)
Data Source
DeepReport Intelligence Briefing — 2026-10-02, sourced from Claude Code User Documentation Review #65018 ("Major Obstacles" — example/smoke-test parity gap), a pattern flagged in every run since at least 2026-08-22 but not previously filed as a standalone issue.
Generated by 🔬 Deep Report · claude · agent · 212 AIC · ⌖ 7.99 AIC · ⊞ 7.3K · ◷
Description
engine: claude(or nestedid: claude) appears 58 times across.github/workflows/*.mdversus 115 for Copilot (~1.98x), and Claude-specific smoke-test variants number only 9 (smoke-claude*.md,smoke-github-claude.md,smoke-claude-on-copilot.md) versus 15 Copilot variants (arm64, SDK, sub-agents, AOAI apikey/Entra auth modes, small-model, etc). A Claude-only user has roughly half the reference material to copy from when building a new workflow, and several auth/runtime modes Copilot covers (sub-agents, ARM64, alternate auth) have no Claude equivalent smoke test.Expected Impact
Closes the reference-material gap for Claude users and gives CI equivalent coverage of Claude-specific runtime modes (ARM64, SDK-style invocation, sub-agents, alternate auth) that currently only Copilot smoke tests exercise.
Suggested Agent
GitHub Copilot coding agent or a workflow-authoring agent — add 4-6 new
smoke-claude-*.mdworkflows mirroring the Copilot variants (e.g.smoke-claude-arm64.md,smoke-claude-sdk.md,smoke-claude-sub-agents.md, an alternate-auth-mode variant), following the existingsmoke-claude.md/smoke-copilot-*.mdstructure, then runmake recompile.Estimated Effort
Medium (1-4 hours)
Data Source
DeepReport Intelligence Briefing — 2026-10-02, sourced from Claude Code User Documentation Review #65018 ("Major Obstacles" — example/smoke-test parity gap), a pattern flagged in every run since at least 2026-08-22 but not previously filed as a standalone issue.