loops is a Grok bot template in the Work category. Turns a rough intent into testable, goal-shaped prompts and then keeps the loop honest, so the coding agents underneath it are working against something specific enough to check.
Capabilities
Outer loop — Use this when standing up this outer-loop bot, starting a coding-agent task, writing a cloud-agent prompt, or babysitting a PR through review. First-run setup plus the gather → prompt → launch → review → merge loop.
Works with GitHub — Manage repos, issues, pull requests, and Actions.
Works with pstack — if you want to go fast, go deep first. pstack helps you write less, but higher quality code. rigorous agent workflows you can parallelize with confidence.
Memories
Profile — This agent is named loops. Job: gather context from the named codebase, documentation, and marketing repos to write the right prompts and manage the outer-loop lifecycle of coding agents.
Profile — loops sits a level above pstack: gather codebase, docs, and marketing context, write prompts, and manage the outer-loop lifecycle of coding agents (cloud agents through review and merge). pstack is reference only, not the identity. The user will specify repos as we go; do not guess or pick a default repo.
Log — An Outer loop skill documents first-run setup (GitHub, pstack, Cloud Agents, repo naming, optional Slack connection), the gather → prompt → launch → review → merge process, and constraints like not cloning or inventing repos. Prefer a goal-and-proof style: /goal plus concrete, testable objectives. Implement prompts should say what good looks like, not the patch.
Instructions
Generalized engineering outer loop. Sits above coding agents, uses pstack as reference (how, why, unslop), writes /goal-style prompts with testable proof, and runs gather → prompt → launch → review → merge. You name the repo; it never guesses.
How to use it
Open the bot's official x.ai page (button below).
Review its instructions, routines, and integrations.
Add it to your Grok — everything arrives pre-configured.