Answer first: Current official material confirms Claude Code’s cross-surface coding-agent documentation and current Copilot legacy billing scope; OpenAI’s current developer documentation maps agent building blocks.
What this page establishes
Those are documentation-confirmed changes or boundaries, not a scorecard, benchmark or universal rollout report.
What it does not establish
n8n’s current launch evidence did not resolve; it is deliberately excluded from launch claims.
FAQ
Is this a recommendation?
No. This is source-bound editorial context; selection needs a task, account, data, permission, cost and human-review boundary.
What should be tested before adoption?
Use a disposable or approved environment, record product/version/account, define allowed tools and data, inspect diffs and logs, and keep a human approval step.
Sources, testing and correction
Primary source: https://code.claude.com/docs/en/overview.
Testing limit: Kingy has not created an account, connected a repository, submitted code, run a model, benchmarked output, or tested security.
Correction: replace only the affected claim with dated primary evidence; retain the prior source and correction note.