dbxj 0.1.0__tar.gz

dbxj-0.1.0/.github/workflows/ci.yml

@@ -0,0 +1,23 @@
+name: CI
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+
+jobs:
+  check:
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        python-version: ["3.11", "3.13"]
+    steps:
+      - uses: actions/checkout@v4
+      - uses: astral-sh/setup-uv@v6
+        with:
+          python-version: ${{ matrix.python-version }}
+      - run: uv sync --locked
+      - run: uv run ruff check
+      - run: uv run ruff format --check
+      - run: uv run mypy src
+      - run: uv run pytest

dbxj-0.1.0/.github/workflows/release.yml

@@ -0,0 +1,17 @@
+name: Release
+
+on:
+  push:
+    tags: ["v*"]
+
+jobs:
+  publish:
+    runs-on: ubuntu-latest
+    environment: pypi
+    permissions:
+      id-token: write  # PyPI trusted publishing
+    steps:
+      - uses: actions/checkout@v4
+      - uses: astral-sh/setup-uv@v6
+      - run: uv build
+      - uses: pypa/gh-action-pypi-publish@release/v1

dbxj-0.1.0/.gitignore

@@ -0,0 +1,180 @@
+### Generated by gibo (https://github.com/simonwhitaker/gibo)
+### https://raw.github.com/github/gitignore/6eeebe6f49678aacd8311ce079842c971b3ebe96/Python.gitignore
+
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# UV
+#   Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#uv.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#pdm.lock
+#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
+#   in version control.
+#   https://pdm.fming.dev/latest/usage/project/#working-with-version-control
+.pdm.toml
+.pdm-python
+.pdm-build/
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+#.idea/
+
+# PyPI configuration file
+.pypirc
+
+# dev
+tmp
+.python-version
+.DS_Store
+cached

dbxj-0.1.0/.justfile

@@ -0,0 +1,29 @@
+# just run <profile>, or just run for the default profile
+run profile="":
+    uv run dbxj {{ if profile != "" { "--profile " + profile } else { "" } }}
+
+run_offline:
+    uv run dbxj --offline
+
+setup_python:
+    uv venv --python 3.11
+    uv sync
+
+upgrade_and_sync:
+    uv upgrade
+    uv sync
+
+sync: # adds all packages to .venv
+    uv sync
+
+create_jobs:
+    uv run ./scripts/create_jobs.py --count 10
+
+create_jobs_50:
+    uv run ./scripts/create_jobs.py --count 50
+
+delete_jobs:
+    uv run ./scripts/delete_jobs.py
+
+install:
+    uv tool install --reinstall .

dbxj-0.1.0/.pre-commit-config.yaml

@@ -0,0 +1,33 @@
+repos:
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v5.0.0
+    hooks:
+      - id: trailing-whitespace
+      - id: end-of-file-fixer
+      - id: check-docstring-first
+      - id: debug-statements
+      - id: check-yaml
+      - id: check-ast
+      - id: check-json
+      - id: pretty-format-json
+        args: [--autofix]
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    # Ruff version.
+    rev: v0.14.0
+    hooks:
+      # Run the linter.
+      - id: ruff
+        args: [--fix]
+      # Run the formatter.
+      - id: ruff-format
+  - repo: local
+    hooks:
+      - id: mypy
+        name: mypy
+        entry: mypy
+        language: system
+        types: [python]
+        args:
+          - --install-types
+          - --non-interactive
+exclude: 'scripts/.*'

dbxj-0.1.0/CHANGELOG.md

@@ -0,0 +1,12 @@
+# Changelog
+
+## 0.1.0
+
+First public release.
+
+- Browse jobs, their runs, the tasks in a run and task output/error traces
+- `Last 5` column with the results of each job's recent runs
+- Filter jobs by name, trigger a job, open the selection in the browser
+- Profile selection (`--profile` / `DATABRICKS_CONFIG_PROFILE`) and in-app re-login when the
+  session has expired
+- `--offline` / `--demo` mode with bundled demo data

dbxj-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024-2026 Travis Dent
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

dbxj-0.1.0/PKG-INFO

@@ -0,0 +1,101 @@
+Metadata-Version: 2.5
+Name: dbxj
+Version: 0.1.0
+Summary: A keyboard-driven terminal UI for browsing Databricks jobs, runs and task logs
+Project-URL: Homepage, https://github.com/pugillum/dbxj
+Project-URL: Issues, https://github.com/pugillum/dbxj/issues
+Author: Travis Dent
+License-Expression: MIT
+License-File: LICENSE
+Keywords: cli,databricks,jobs,textual,tui
+Classifier: Development Status :: 4 - Beta
+Classifier: Environment :: Console
+Classifier: Intended Audience :: Developers
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Software Development
+Classifier: Topic :: Utilities
+Requires-Python: >=3.11
+Requires-Dist: databricks-sdk>=0.41.0
+Requires-Dist: textual>=1.0.0
+Description-Content-Type: text/markdown
+
+# dbxj
+
+A keyboard-driven terminal UI for browsing Databricks jobs, their runs and task logs, in the
+spirit of k9s and lazygit.
+
+![dbxj showing a list of jobs with their last 5 run results](https://raw.githubusercontent.com/pugillum/dbxj/main/docs/screenshot.svg)
+
+- See every job with the results of its last 5 runs at a glance
+- Drill into a job's runs, a run's tasks and a task's output or error trace
+- Filter jobs by name, trigger a job, open anything in the browser
+- Switch workspaces with `--profile`; re-login happens in-app when your session expires
+
+# Install
+
+```sh
+uv tool install dbxj      # or: pipx install dbxj
+```
+
+Requirements:
+
+- Python 3.11+
+- A profile in `~/.databrickscfg`, e.g. from `databricks auth login --host <workspace-url>`
+- The [Databricks CLI](https://docs.databricks.com/dev-tools/cli/install.html) on your `PATH`, used
+  to log in again when your session has expired
+
+Try it without a workspace:
+
+```sh
+dbxj --demo
+```
+
+Upgrade with `uv tool upgrade dbxj` and remove with `uv tool uninstall dbxj`.
+
+# Usage
+
+```sh
+dbxj --profile <profile name>   # a profile from ~/.databrickscfg
+dbxj --demo                     # bundled demo data, no workspace needed (alias: --offline)
+```
+
+If `--profile` isn't given, `DATABRICKS_CONFIG_PROFILE` or the SDK's default auth is used. On start
+the login is checked; if it's expired a dialog appears and `databricks auth login` opens your browser,
+then the app continues once you've logged in.
+
+| Key | Action |
+| --- | --- |
+| `r` | Refresh the current view |
+| `q` | Quit |
+| `?` | Show all key bindings (`esc` closes) |
+| `/` | Filter jobs by name (kept while you drill in) |
+| `t` | Trigger the selected job (asks to confirm) |
+| `o` | Open the selected job, run or task in the browser |
+| `enter` / `esc` | Drill into / back out of jobs, runs and tasks |
+| `j` / `k`, `d` / `u` | Down / up, page down / up |
+
+# Development
+
+Requires [uv](https://docs.astral.sh/uv/) and optionally [just](https://github.com/casey/just).
+
+```sh
+uv sync                    # create .venv
+uv run pre-commit install
+uv run dbxj --demo         # or: just run_offline / just run <profile>
+uv run pytest
+```
+
+- In demo mode, data is read from `./cached` if that directory exists, otherwise from
+  `src/dbxj/sample_data` (regenerate with `uv run scripts/generate_sample_data.py`)
+- `scripts/` has helpers to create, trigger and delete test jobs in a workspace
+- `docs/design.md` describes how the app is put together
+
+Releases are published to PyPI by pushing a `v*` tag.
+
+# License
+
+MIT

dbxj-0.1.0/README.md

@@ -0,0 +1,76 @@
+# dbxj
+
+A keyboard-driven terminal UI for browsing Databricks jobs, their runs and task logs, in the
+spirit of k9s and lazygit.
+
+![dbxj showing a list of jobs with their last 5 run results](https://raw.githubusercontent.com/pugillum/dbxj/main/docs/screenshot.svg)
+
+- See every job with the results of its last 5 runs at a glance
+- Drill into a job's runs, a run's tasks and a task's output or error trace
+- Filter jobs by name, trigger a job, open anything in the browser
+- Switch workspaces with `--profile`; re-login happens in-app when your session expires
+
+# Install
+
+```sh
+uv tool install dbxj      # or: pipx install dbxj
+```
+
+Requirements:
+
+- Python 3.11+
+- A profile in `~/.databrickscfg`, e.g. from `databricks auth login --host <workspace-url>`
+- The [Databricks CLI](https://docs.databricks.com/dev-tools/cli/install.html) on your `PATH`, used
+  to log in again when your session has expired
+
+Try it without a workspace:
+
+```sh
+dbxj --demo
+```
+
+Upgrade with `uv tool upgrade dbxj` and remove with `uv tool uninstall dbxj`.
+
+# Usage
+
+```sh
+dbxj --profile <profile name>   # a profile from ~/.databrickscfg
+dbxj --demo                     # bundled demo data, no workspace needed (alias: --offline)
+```
+
+If `--profile` isn't given, `DATABRICKS_CONFIG_PROFILE` or the SDK's default auth is used. On start
+the login is checked; if it's expired a dialog appears and `databricks auth login` opens your browser,
+then the app continues once you've logged in.
+
+| Key | Action |
+| --- | --- |
+| `r` | Refresh the current view |
+| `q` | Quit |
+| `?` | Show all key bindings (`esc` closes) |
+| `/` | Filter jobs by name (kept while you drill in) |
+| `t` | Trigger the selected job (asks to confirm) |
+| `o` | Open the selected job, run or task in the browser |
+| `enter` / `esc` | Drill into / back out of jobs, runs and tasks |
+| `j` / `k`, `d` / `u` | Down / up, page down / up |
+
+# Development
+
+Requires [uv](https://docs.astral.sh/uv/) and optionally [just](https://github.com/casey/just).
+
+```sh
+uv sync                    # create .venv
+uv run pre-commit install
+uv run dbxj --demo         # or: just run_offline / just run <profile>
+uv run pytest
+```
+
+- In demo mode, data is read from `./cached` if that directory exists, otherwise from
+  `src/dbxj/sample_data` (regenerate with `uv run scripts/generate_sample_data.py`)
+- `scripts/` has helpers to create, trigger and delete test jobs in a workspace
+- `docs/design.md` describes how the app is put together
+
+Releases are published to PyPI by pushing a `v*` tag.
+
+# License
+
+MIT

dbxj-0.1.0/docs/design.md

@@ -0,0 +1,148 @@
+# Design
+
+`dbxj` is a terminal UI for browsing Databricks jobs, their runs, the tasks in a run and task output.
+It's built on [Textual](https://textual.textualize.io/) and the
+[Databricks Python SDK](https://github.com/databricks/databricks-sdk-py).
+
+## Style: minimal TUI like lazygit and k9s
+
+`dbxj` should look and feel like [lazygit](https://github.com/jesseduffield/lazygit) and
+[k9s](https://k9scli.io/), not like a typical Textual app. Every UI change must follow these rules:
+
+- **Keyboard only.** Every action is a single key (`t`, `o`, `r`, `/`, ...) with vim-style
+  navigation (`j`/`k`, `d`/`u`). There are no buttons, checkboxes or clickable widgets, and nothing
+  that needs the mouse.
+- **Plain bordered panels.** Use `solid white` borders on the default background. Put titles and key
+  hints in the border (`border_title` / `border_subtitle`, e.g. `y confirm · n cancel`), not in extra
+  widgets. Don't add Textual's stock look: no themed dialog boxes, header bars or coloured
+  backgrounds.
+- **Popups are overlays.** Dialogs, confirmations and help are a small centred box on a
+  `ModalScreen` over the current view (see `dialogs.py`). `esc` always closes them. `y`/`n` answers
+  confirmations. Textual's default dimming of the screen behind a modal is turned off
+  (`background: transparent`). The box itself is opaque (`background: $background`), so the table
+  doesn't show through.
+- **Dense, text-first content.** Prefer tables and compact symbols (`✓ ✗ ● ⊘`) with sparing colour
+  (green/red/yellow/grey for state, yellow for keys). Don't use decorative padding or spacing.
+- **One bottom bar.** `StatusBar` (`main.py`) holds the key hints on the left and a brief status
+  message on the right, like lazygit. The hints are only the essential keys (`r` refresh, `q` quit,
+  `?` help). Every other binding uses `show=False` and is listed in the `?` help overlay. The `^p`
+  command palette hint is hidden too, though `ctrl+p` still works.
+- **Feedback is a status message, never an overlay.** `DatabricksTUI.notify` is overridden, so every
+  `notify()` call writes one line into the bottom bar instead of showing Textual's toast cards.
+  Messages are green / yellow / red for information / warning / error and clear themselves after
+  4s, or 8s for errors. Use a popup only when the user has to make a decision.
+- **No scrollbars.** A global `* { scrollbar-size: 0 0; }` in `main.tcss` hides them. In a narrow or
+  short window the content is clipped like in k9s and scrolled with keys (`j`/`k`, `d`/`u`, and
+  `h`/`l` or arrows to pan wide tables).
+- **Restyle stock Textual widgets rather than accepting their defaults.** Check new widgets against
+  these rules. Textual's `DEFAULT_CSS` is overridden by `main.tcss`.
+
+## Principles
+
+- **`q` / `esc` always work**, including while waiting on something slow (e.g. browser login).
+- **Destructive or costly actions ask first**, e.g. `t` triggering a job shows a y/n popup.
+- **Never block the UI thread**: API calls run in Textual workers (`@work(thread=True)` or
+  `asyncio.to_thread`).
+
+## Layout
+
+```
+src/dbxj/
+  main.py                    entry point, arg parsing, App, MainScreen, SplashScreen, Header
+  body.py                    Body widget: table/filter/details panels and the navigation state machine
+  jobs.py                    JobsAPI: cached job list
+  auth.py                    profiles, connection check, ReauthScreen
+  dialogs.py                 HelpScreen (? overlay), ConfirmScreen (y/n popup)
+  mock_workspace_client.py   offline client backed by ./cached/*.json or sample_data/*.json
+  sample_data/               fake workspace for --offline/--demo (scripts/generate_sample_data.py)
+  main.tcss                  styles
+```
+
+## Startup flow
+
+1. `main()` parses `--profile` (default `DATABRICKS_CONFIG_PROFILE`), `--offline` (alias `--demo`)
+   and `--version`. An unknown
+   profile exits with the list from `~/.databrickscfg` (or `DATABRICKS_CONFIG_FILE`).
+2. `DatabricksTUI` pushes `MainScreen`, then `SplashScreen` on top.
+3. `MainScreen.setup_client` (async worker) calls `auth.connect(profile)` in a thread. It builds a
+   `WorkspaceClient` and calls `current_user.me()` / `get_workspace_id()` to prove the login is valid.
+4. On `AuthError` it awaits `ReauthScreen` (see below). If login succeeds it loops back to step 3;
+   if the user quits, the app exits.
+5. `Header` and `Body` are initialised. When the first job list loads, `Body` posts
+   `JobsInitialised` and the app pops the splash screen.
+
+Offline mode skips 3–4 and uses `get_mock_client()`. It reads `./cached` if that directory exists
+(for local development against captured data), otherwise the fake workspace bundled in
+`dbxj/sample_data`, so `dbxj --demo` works after a plain install. `tests/test_smoke.py` drives the
+app against that data.
+
+## Re-authentication
+
+All profiles are expected to use `auth_type = databricks-cli` (OAuth U2M), so re-auth shells out to
+`databricks auth login --profile <p>`. The CLI opens the browser and blocks until the OAuth callback.
+
+`ReauthScreen` (a `ModalScreen[bool]`):
+- starts the login as soon as it mounts, with output captured so it can't corrupt the TUI
+- `q` / `esc` terminates the login process and dismisses with `False`
+- on failure it shows stderr; `enter` / `r` retries
+- any exception from `connect` is treated as an auth failure; the error text is shown so network
+  problems can be told apart
+
+## Navigation state machine (`body.py`)
+
+`BodyState`: `JOBS` → `JOB_RUNS` → `JOB_RUN` (tasks) → `TASK` (logs). `enter` drills in, `esc` goes
+back. Each level clears the shared `DataTablePlus` and reloads it in a worker. `TASK` hides the table
+and shows `DetailsPanel`, a scrollable view of the run output with ANSI colours.
+
+`FilterState`: `NONE` → `TYPING` (`/` opens the filter input; the table filters live) → `SELECTING`
+(`enter` moves focus back to the table and saves the text in `Body.job_filter`). Only available in
+`JOBS`. Drilling into a job only hides the filter panel; `esc` back to `JOBS` shows it again and
+reapplies `job_filter`. `esc` in `JOBS` clears the filter.
+
+Tables key their rows by job / run id. Going back with `esc` puts the cursor back on the row that
+was drilled into (`current_job_key`, `current_run_key`, `current_task_run_id`), via the loaders'
+`select_key` argument.
+
+## Actions
+
+Bound on `MainScreen` and delegated to `Body`:
+
+- `r` → `Body.refresh_view`: reloads the current view (jobs with the cache cleared and the filter
+  kept, runs, tasks, or the task logs), keeping the cursor on the same row.
+- `t` → `Body.trigger_job`: jobs view only. It shows a `ConfirmScreen`, then `jobs.run_now` in a
+  worker. It doesn't wait on the returned `Wait`; the new run id comes from `.bind()["run_id"]`. A
+  status message in the bottom bar reports the result, and the job's `Last 5` cell is refetched.
+- `o` → `Body.open_in_browser` → `App.open_url`. Jobs use `{host}/jobs/{id}?o={workspace_id}`. Runs
+  and tasks use the API's `run_page_url`, with a fallback of `{host}/jobs/{job}/runs/{run}?o=...`.
+- `?` → `HelpScreen`, a static key list in `dialogs.HELP_SECTIONS`. Most navigation keys are
+  handled in `Body.on_key` rather than as `Binding`s, so they can't be generated.
+
+Single-letter screen bindings don't fire while typing in the filter, because `Input` consumes
+printable keys.
+
+## Last 5 runs column
+
+The jobs table has a `Last 5` column, ordered oldest → newest: `✓` success, `✗` failed, `●` running/
+queued, `⊘` cancelled/skipped, `-` no run, `?` fetch failed, `…` loading.
+
+The public Jobs API can't return recent runs for many jobs in one call. `jobs.list` has no runs
+option and `jobs.list_runs` takes a single `job_id`. The portal uses an internal GraphQL route
+(`/graphql/LakeflowListAssetsQuery`) that only accepts browser session cookies: bearer tokens get a
+303 to the login page (tested 2026-11-14), so we can't use it. Instead `Body.load_recent_runs` makes one
+`list_runs(job_id, limit=5)` per job on an 8-thread pool. The table renders at once and cells fill in
+as results arrive. The SDK iterator follows `next_page_token`, so `JobsAPI.get_recent_runs` caps the
+results with `islice`. The column has a fixed width (`2 * RECENT_RUNS - 1`): `update_cell` doesn't resize a column, so
+sizing it from the `…` placeholders cut cells off at 3 symbols. Results (including failures) are cached in `JobsAPI.recent_runs`, so filtering
+doesn't refetch. `r` clears the cache.
+
+## Packaging
+
+- The package and command are both `dbxj` (`[project.scripts] dbxj = "dbxj.main:main"`), installed
+  with `uv tool install .`.
+- Runtime deps are only `databricks-sdk` and `textual`; tooling lives in the `dev` group.
+- `just run [profile]`, `just run_offline`, `just install`.
+
+## Known gaps
+
+- Offline mode needs a `./cached/` directory of JSON fixtures that isn't in the repo.
+- Run/task lists are capped at 25 and not paginated.