Your output won't scale
if you're in every
little loop
little-loops imagines Software Engineering as the art of crafting simple machines that run themselves. A loop to discover bugs. A loop to score results. A loop to write code that feels inevitable. A loop made of these loops.
/plugin marketplace add BrennonTWilliams/little-loops
/plugin install ll@little-loopsYou were promised
exponential output.
You got
babysitting chat windows.
What if...
// designed to run without you
Wind it up. Let it run.
A loop is a small machine you define once and run whenever you need it. Describe what to check, what to fix, and when to stop. little-loops compiles it into a finite state machine and hands it to Claude. No babysitting. No streaming output to watch. Just a goal and a max iteration count.
Never write a disposable prompt again.
Start from a conversation, a pattern you keep repeating, or a blank slate.
/ll:create-loopInteractive wizard — answer a few questions, get a YAML loop file.
/ll:loop-suggesterMines your session history for repeated workflows and proposes loops that would automate them.
Resilience
State saves to disk after every FSM transition. Crashes, context resets, Ctrl-C — ll-loop resume picks up exactly where it stopped.
Coordination
Agents automatically hand off to new agents when context gets long. Context rot is now a 2025 problem.
Context that travels with your code.
Issues in little-loops are mobile encapsulations of a unit of work — tracking the thinking, planning, decisions, and artifacts. Then on completion, they become the permanent memory of your project.
Find what's broken. Capture what you notice. Standardize everything.
You can't fix what you haven't found. little-loops finds it — surfacing bugs, architectural gaps, and dead code. Capture anything you spot mid-conversation.
/ll:scan-codebase # technical issues: bugs, dead code, type gaps
/ll:audit-architecture # structural problems: coupling, dependency cycles
/ll:capture-issue # capture anything spotted mid-conversation
/ll:normalize-issues # standardize format across all captured issuesEvery issue becomes a structured Markdown file in .issues/ — not a note, not a comment. A specification.
One issue or thirty. Sequential or parallel. Your call.
Once your issues are refined and ready, hand them off. Choose sequential for simplicity, parallel for speed, or sprint for coordinated wave execution across complex dependency graphs.
Sequential. Predictable. Fire and forget.
Picks up every ready issue in priority order and works through them one at a time. Deterministic execution — each issue completes before the next begins. Great for auditable, reviewable batch work.
ll-auto
ll-auto --max-issues 5
ll-auto --resumeDiscover. Refine. Execute. Learn. Repeat.
| # | Step | Command | What happens |
|---|---|---|---|
| 1 | Discover | /ll:scan-codebase | Surface bugs, gaps, architectural problems |
| 2 | Prune | /ll:verify-issues + /ll:tradeoff-review-issues | Kill stale issues and low-value noise |
| 3 | Refine | /ll:refine-issue + /ll:ready-issue | Fill gaps, gate quality, greenlight implementation |
| 4 | Execute | ll-sprint run sprint-1 | Parallel waves across isolated git worktrees |
| 5 | Validate | ll-loop run quality-gate | Quality loop runs until all gates green |
| 6 | Ship | /ll:commit + /ll:open-pr | Scoped commits, generated PR descriptions |
| 7 | Learn | /ll:analyze-workflows | Mine your history — turn repeated patterns into new loops |
Use issues, sprints, and loops independently or chain them. The same workflow scales from a single bug fix to a week of parallel feature work.
# 1. Add the plugin marketplace
claude plugin marketplace add BrennonTWilliams/little-loops
# 2. Install the plugin
claude plugin install ll@little-loops
# 3. Install the CLI tools via pip
pip install little-loops# 4. Initialize for your project
/ll:init
# 5. Start scaling
/ll:audit-claude-config
/ll:scan-codebase