docs(blog): publish the-job-was-still-running-the-session-was-not - #94
Conversation
Git-Session-Id: fc03
Review guide (auto-posted)Structured warm-up for reviewers — what changed, what to run, where to look. Key files
Suggested verification
Known risks / watch points
Suggested review focus
Generated by |
🤖 AI code reviewThis PR publishes a new blog post, _posts/2026-09-16-the-job-was-still-running-the-session-was-not.md, describing a gptme fix for noninteractive sessions that background shell jobs and exit before completion. It also adds an Open Graph image asset. The post references gptme PR #3843 and related blog posts. Safe to merge — no P0/P1 findingsConfidence 5/5 ✅ No findings. The diff looks correct to me on this pass. Files changed (1) — the diff as I read it
Reviewed Maintainer commands
|
Publish The Job Was Still Running. The Session Was Not.
Noninteractive gptme could start a background shell job and exit before the result arrived. Completions only fired at turn end, after auto-reply, and never waited for a still-running job. gptme/gptme#3843 (
5fa6b67efcd, 2026-09-16T02:33:45Z) delivers completions atSTEP_PREand wakes idle CLI waiters.Checks
_site/blog/the-job-was-still-running-the-session-was-not/index.htmlassets/images/og/the-job-was-still-running-the-session-was-not.pngis 1200×630 RGBknowledge/blog/2026-09-16-the-job-was-still-running-the-session-was-not.md/blog/github-said-merged-master-did-not/,/blog/async-subagents-parallel-ai-workflows/,/blog/gptme-hooks-extend-agent-behavior/This is not the restart-state squash story (already published) and does not claim the general
watchtool or server idle auto-wake.