looper
Design visual, review-gated agent loops for Claude Code before you run them.
What is it?
What it is
Looper is a loop design coach for Claude Code that helps users create well-structured agent loops with clear goals, checkable verification, and review gates before execution.
Why it exists
To address gaps in existing loop commands by enforcing design discipline, ensuring loops are reviewed and validated against best practices before running, and enabling cross-model reviewers or judges.
Who should use it
Developers working on agentic workflowsAI/ML engineers designing automated loopsDevOps engineers needing structured verificationProject managers requiring reviewable loop specs
Who should avoid it
Users without Claude Code integrationTeams needing simple one-off automation without design reviewBeginners unfamiliar with agent loops or YAML
How it works
A quick walkthrough in plain English
How looper works
Step 1 of 3
You interact with it
Open looper, send a request, or connect it to your stack.
Features
Advantages
- Ensures loops are well‑designed and reviewable before running
- Provides clear, checkable success criteria and termination conditions
- Facilitates cross‑vendor review to reduce blind spots
- Produces portable artifacts that can be versioned and reused
- Offers static linting to catch design anti‑patterns early
- Redaction safeguards sensitive data automatically
- Supports both quick in‑session runs and repeatable external execution
- Encourages best practices through coaching and templates
- Simplifies integration with Claude Code tooling and workflows
- Allows custom runners to be verified against a contract
Disadvantages
- Does not provide durable orchestration or scheduling beyond Claude Code’s own tools
- Requires installation of the skill and command in Claude Code environment
- Python runner and spec are Python‑centric; other languages need custom implementation
- Learning curve for understanding the design rubrics and loop structure
- Redaction may over‑mask useful content if globs are too broad
- Limited to the features of the underlying LLM APIs and CLI tooling
- Cannot enforce concurrency or persistent state across restarts without external orchestrator
Installation
native
On Windows PowerShell: irm https://raw.githubusercontent.com/ksimback/looper/main/install.ps1 | iex On macOS/Linux: curl -fsSL https://raw.githubusercontent.com/ksimback/looper/main/install.sh | bash Manual install commands: Windows PowerShell: git clone https://github.com/ksimback/looper "$env:USERPROFILE\.claude\skills\looper" New-Item -ItemType Directory -Force "$env:USERPROFILE\.claude\commands" | Out-Null Copy-Item "$env:USERPROFILE\.claude\skills\looper\commands\looper.md" "$env:USERPROFILE\.claude\commands\looper.md" -Force macOS/Linux: git clone https://github.com/ksimback/looper "$HOME/.claude/skills/looper" mkdir -p "$HOME/.claude/commands" cp "$HOME/.claude/skills/looper/commands/looper.md" "$HOME/.claude/commands/looper.md" Uninstall: Windows PowerShell: Remove-Item -Recurse -Force "$env:USERPROFILE\.claude\skills\looper" Remove-Item -Force "$env:USERPROFILE\.claude\commands\looper.md" Remove-Item -Recurse -Force "$env:USERPROFILE\.looper" macOS/Linux: rm -rf "$HOME/.claude/skills/looper" rm -f "$HOME/.claude/commands/looper.md" rm -rf "$HOME/.looper"
FAQ
How do I install Looper?
On Windows PowerShell: `irm https://raw.githubusercontent.com/ksimback/looper/main/install.ps1 | iex`. On macOS/Linux: `curl -fsSL https://raw.githubusercontent.com/ksimback/looper/main/install.sh | bash`.
What templates does Looper provide?
Looper offers templates like `security-scan`, `code-review`, `bug-hunt`, `docs-sync`, and `research-synthesis` for common loop use cases.
Can I run a loop outside the current session?
Yes, use `python3 ./looper-output/run-loop.py` to execute the loop externally after Looper generates the spec.
How does Looper handle sensitive data?
Looper uses two-layer redaction: files matching redaction globs are never read into prompts, and content from those files is scrubbed before every send, with logs in `run-log.md`.
What files does Looper create?
Looper generates `loop.yaml`, `loop.resolved.json`, `RUN_IN_SESSION.md`, `run-loop.py`, an empty `loop-workspace/`, and a README for the loop.
How do I validate a loop design?
Use `python scripts/looper.py lint path/to/loop.yaml` to check for design anti-patterns and ensure the spec is compiler-valid.
Featured in Videos
YouTube tutorials and walkthroughs for looper
Alternatives
Similar projects ranked by category, topics, and text overlap.
