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.

These functions manage the IssueLoop ticket store over time — pruning old resolved tickets, recovering stale in-progress claims, exporting data, and rotating large log files. The default backend is a local SQLite file at data/issueloop.db. All functions that modify the database commit immediately.

​
cleanup

Deletes tickets that have been in a terminal status (done or failed by default) for longer than the retention period. This is the primary housekeeping function for long-running agent deployments. 
removed = issueloop.cleanup()
removed = issueloop.cleanup(older_than_days=7, repo="myrepo", statuses=["done"])
​
older_than_days
int
Delete tickets older than this many days. Defaults to the retention_days value from config (default 30).
​
repo
str
Optional repository filter. None cleans up tickets across all repos.
​
statuses
list
List of status strings to consider for deletion. Defaults to ["done", "failed"]. Only tickets in one of these statuses are eligible.
Returns an int — the number of tickets that were deleted.

​
purge_repo

Deletes all tickets for a repository, regardless of status or age. Use with care — this is irreversible. 
removed = issueloop.purge_repo("myrepo")
​
repo
str
required
Name of the repository whose tickets should be deleted.
Returns an int — the number of tickets deleted.

​
get_database_stats

Returns a summary of the ticket store, including per-status counts and (for the local SQLite backend) the file size on disk. 
stats = issueloop.get_database_stats()
stats_for_repo = issueloop.get_database_stats(repo="myrepo")
​
repo
str
Optional repository filter. None returns stats across all repos.
Returns a dict with:
​
database
str
The configured backend type, e.g. "local".
​
total_bugs
int
Total ticket count in the filtered scope.
​
by_status
dict
Mapping of status string to ticket count. Only statuses with at least one ticket appear.
​
db_size_bytes
int
SQLite file size in bytes. Present only when using the "local" backend.

​
export_bugs

Exports all tickets (or tickets for one repo) to a JSON file and returns the path to that file. 
path = issueloop.export_bugs()
path = issueloop.export_bugs(repo="myrepo", path="/tmp/myrepo_bugs.json")
​
repo
str
Optional repository filter. None exports all tickets.
​
path
str
Output file path. Defaults to data/logs/bug_export.json. Parent directories are created if they do not exist.
Returns a str — the absolute path of the file that was written. The file contains a JSON array of ticket dicts (same shape as the Ticket object).

​
reap_stale_bugs

Finds tickets that have been in in_progress status for longer than older_than_minutes minutes and resets them to pending. This handles the case where a worker process claimed a ticket and then crashed or was killed before finishing. 
reaped = issueloop.reap_stale_bugs()
reaped = issueloop.reap_stale_bugs(older_than_minutes=60, repo="myrepo")
​
older_than_minutes
int
default:"30"
Tickets that have had dispensed_at set for longer than this many minutes are eligible for reaping.
​
repo
str
Optional repository filter.
Returns an int — the number of tickets reset to pending.

​
rotate_logs

Archives the JSONL run log for one or all repositories when it exceeds the size threshold. The active log is renamed to run_<repo>.YYYYMMDDTHHMMSSZ.jsonl.bak and the read cursor is reset so the next create_tickets call starts fresh. 
# Rotate logs for all repos (only archives if above threshold)
results = issueloop.rotate_logs()

# Rotate one repo with a custom size limit
results = issueloop.rotate_logs(repo_name="myrepo", max_size_mb=5.0)
​
repo_name
str
Optional repository name. When provided, only that repo’s log is checked. When omitted, all repos in the manifest are checked.
​
max_size_mb
float
default:"10.0"
Size threshold in megabytes. A log file must exceed this size to be archived. Files at or below the threshold are left untouched.
Returns a dict mapping repo name to True (archived) or False (no rotation needed).
Schedule cleanup() and reap_stale_bugs() on a regular interval in long-running agent loops. A cron-style wrapper that calls these every hour prevents the SQLite file from growing indefinitely and ensures crashed workers do not permanently hold tickets.

Build docs developers (and LLMs) love

Get started for freeTalk to us