Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/onenot8/issueLoop/llms.txt

Use this file to discover all available pages before exploring further.

The issueloop CLI is the primary interface for running IssueLoop without writing Python code. It wraps the Python API into shell-friendly subcommands that map one-to-one onto the underlying library functions — so everything you can do from a Python script you can also drive entirely from a terminal, a Makefile, or a CI pipeline.

Available commands

The table below lists every subcommand and what it does at a glance. Full argument references and output examples are in the Commands page.
CommandDescription
check-envVerify Ollama, models, and Supabase env are ready
scan <repo_path>Build a file inventory for a repo
test <repo_name>Run test_manifest.json commands, write log cache
tickets <repo_name>Create tickets from failing log entries
next <repo_name>Dispense the next ticket, one at a time
watch <repo_name>Watch a live process or log file for errors
serveStart the local HTTP bridge
cleanupDelete old done/failed tickets

Global usage

Every issueloop invocation accepts a --help flag. Use the top-level flag to see the list of subcommands, and the per-command flag to see that command’s arguments:
issueloop --help
issueloop <command> --help
For example:
issueloop watch --help

Typical workflow

The commands are designed to run in sequence. Each step feeds the next: scan writes a file inventory, test reads it and writes log output, tickets reads the logs and creates tickets, and next claims them one at a time.
# 1. Confirm your environment is configured correctly
issueloop check-env

# 2. Build a file inventory for the repo you want to monitor
issueloop scan repos/myrepo

# 3. Run the test suite and write the log cache
issueloop test myrepo

# 4. Ask the LLM to split failures into individual tickets
issueloop tickets myrepo

# 5. Claim the highest-priority ticket and start working on it
issueloop next myrepo

# 6. (Optional) Watch a live process for runtime errors
issueloop watch myrepo --command "python3 main.py" --cwd repos/myrepo

# 7. (Optional) Expose the ticket queue over HTTP for non-Python callers
issueloop serve --port 8787

# 8. Periodically remove tickets that are resolved or failed
issueloop cleanup --days 30
Each CLI command calls the equivalent Python function directly. issueloop scan repos/myrepo calls issueloop.scan_repo("repos/myrepo"), issueloop next myrepo calls issueloop.get_top_error("myrepo"), and so on. If you need programmatic control — looping, branching, or combining calls — switch to the Python API instead of shelling out.

Build docs developers (and LLMs) love