Editorial picks
Best coding Grok bots.
Grok Bot is the first joint xAI–Cursor product and runs on Cursor infrastructure, so coding is its home turf. These five templates cover the loop end to end: specs grounded in shipped work, a full-stack builder, a night shift that queues pull requests, a reviewer that has to be convinced, and a research bot for ramping onto new territory.
Check what each one expects before adding it — codebase, shell, browser, or deployment access. Read-only reviewers are low-risk; anything that can push, deploy, or spend should stay behind approvals until you trust the workflow.
Last reviewed
5 templates · every pick argued for
Browse all templates →A full-stack product builder with enough interface taste to move beyond scaffolding.
Why it made the listThe benchmark coding template here: a rough brief in, and planning, implementation, debugging, and launch out, with enough interface taste that the result moves beyond scaffolding. Review its write and deploy permissions before letting it ship.
EngineeringOn demandShared by @TomZarebczan · link checked 29 Aug 2026Stands between a working branch and your main line as the reviewer who has to be convinced. It reads what a change genuinely alters in the code, sets that against the summary its aut…
Why it made the listA reviewer that has to be convinced: it reads what a change genuinely alters in the code, sets that against the summary its author wrote, and calls out any distance between the two. Nothing gets signed off until the suite has finished green.
EngineeringOn demandShared by @inqusit · link checked 30 Aug 2026Night-shift coding bot: owner queues an idea; it ships PRs to review in the morning.
Why it made the listA night shift for one product: the owner queues an idea and finds pull requests ready to review in the morning. Because the output is a PR rather than a deploy, every change still passes through a human first.
EngineeringLikely scheduledShared by @joshkimWrite grounded product specs and discovery docs directly into ChatPRD.
Why it made the listThe upstream half of coding work: it turns meeting notes, existing documents, and shipped issue history into product specs and discovery docs, so the brief you hand a coding bot is grounded in what the team actually built.
AssistantsOn demandShared by @clairevo · link checked 29 Aug 2026A filtered learning stack drawn from papers, code, lectures, forums, and primary docs.
Why it made the listFor the ramp-up half of engineering: name one AI topic and it filters papers, GitHub, lectures, courses, forums, and lab docs into a ranked learning path. Basic use needs nothing beyond public web access, so it is a low-risk first add.
ResearchOn demandShared by @beamnxw · link checked 29 Aug 2026