Changelog¶

v0.4.0¶

New entry point — The library is now accessed through a single JiraAPI facade. The previous per-class imports (from jira2py import Issues) have been removed. See Installation for the new usage.
Constructor parameter names changed:

v0.3.x v0.4.0

jira_url url

jira_user username

jira_api_token api_token
raw_response mode removed — All methods return parsed JSON (dict, list, or None).
Method signature changes — Several methods had parameters removed in favor of extra_params / extra_data. See API Reference — Conventions for how to pass additional parameters.
- get_issue() — removed: fields_by_keys, properties, update_history, fail_fast
- edit_issue() — removed: override_screen_security, override_editable_flag, history_metadata, properties, transitions, update, additional_properties
- enhanced_search() — removed: properties, fields_by_keys, fail_fast, reconcile_issues
Return type changes:
- edit_issue() returns None on success instead of True. Returns dict when return_issue=True.
- get_changelogs() returns the full paginated response dict instead of just the values list.
Error types changed — All errors were previously raised as ValueError. They are now specific exception types. See Error Handling.
Pydantic validation removed — @validate_call decorators are no longer used. Type hints remain for IDE support and static analysis.

Unified API facade — Single JiraAPI entry point with access to all modules via properties (jira.issues, jira.search, jira.comments, etc.).
Automatic rate limit handling — Requests that receive HTTP 429 are retried automatically with exponential backoff and Retry-After header support. See Rate Limiting.
Structured exception hierarchy — Typed exceptions for authentication errors, not found, validation failures, rate limits, and connection issues. See Error Handling.

11 new API methods:

Method	Description
`issues.create_issue()`	Create a new issue
`issues.get_edit_metadata()`	Get fields available for editing
`issues.get_create_issue_types()`	Get issue types for a project
`issues.get_create_fields()`	Get fields for creating an issue type
`comments.add_comment()`	Add a comment to an issue
`issue_links.get_link_types()`	List available link types
`issue_links.create_link()`	Link two issues
`issue_links.delete_link()`	Delete an issue link
`attachments.get_attachment_metadata()`	Get attachment metadata
`projects.search_projects()`	Search and list projects
`users.search_users()`	Search users by name or email

HTTP/2 support with persistent connections and configurable timeouts.
PEP 561 compliant — py.typed marker included for downstream type checking.

get_comments() ordering — Fixed the order_by query parameter name from orderby to orderBy. The incorrect name was silently ignored by Jira, so comments were always returned in default order.

Migrated from Sphinx to MkDocs Material — complete rewrite of all documentation with new content, Material theme (light/dark toggle, navigation tabs, search, code copy, content tabs).
Machine-readable documentation for AI agents:
- llms.txt / llms-full.txt — following the llmstxt.org standard, generated by mkdocs-llmstxt
- api-reference.json — full API schema with signatures, types, and docstrings, generated from source code by griffe
Documentation dependency changes:

Removed Added

sphinx mkdocs-material

pydata-sphinx-theme mkdocs

sphinx-sitemap mkdocs-llmstxt

griffe

Initial public release.