// the find
ansible-community/ara
ARA Records Ansible and makes it easier to understand and troubleshoot.
ARA is an Ansible callback plugin + Django REST API + web UI that records every playbook run to SQLite, MySQL, or PostgreSQL and lets you browse results, filter by host/task/status, and search across runs. It's for teams running Ansible regularly who've outgrown reading stdout and want a persistent, queryable history of what ran, on what, and what failed. Works without a server (local SQLite default) or with a shared central server for aggregating runs from CI and multiple machines.
The zero-config path is genuinely good — one pip install, one env var, and every subsequent ansible-playbook run is recorded automatically with no playbook changes required. The distributed SQLite backend is a clever design: each playbook writes its own SQLite file, and a custom Django database backend stitches them together at query time, so you get a shared UI without needing a central server during execution. The MCP server in contrib/ means you can point an LLM at your playbook history and ask it questions, which is a practical use of AI for ops debugging rather than a gimmick. Test coverage is thorough — separate test files per resource type, integration tests against real databases, and Zuul CI configs that actually run the integration suite.
The web UI is read-only Django templates with Bootstrap — fine for browsing but there's no way to re-run a failed playbook, compare two runs diff-style, or annotate a failure from the UI. The data model has no concept of environments or namespaces, so if you run the same playbook against staging and production, they show up as two undifferentiated entries that you have to manually distinguish by label. Authentication is an afterthought — the docs note it's disabled by default and recommend enabling it 'for production use,' but token-based auth with no RBAC means everyone with a token sees everything. The project migrated from GitHub to Codeberg but still has GitHub issue templates and references in docs, which signals some organizational churn; commit velocity has slowed noticeably.