Complete Textual observation and execution workflows for Auton 1.3.0

.github/workflows/tests.yml

@@ -5,6 +5,36 @@ on:
 permissions:
   contents: read
 jobs:
+  textual:
+    name: Optional Textual candidate
+    runs-on: ubuntu-latest
+    timeout-minutes: 15
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/checkout@v4
+        with:
+          repository: decryptus/dwho
+          ref: 50b50eaa9696628e9521a31ae92305e01966e6fa
+          path: candidate-dwho
+          persist-credentials: false
+      - uses: actions/setup-python@v5
+        with:
+          python-version: '3.12'
+      - run: python -m pip install ./candidate-dwho
+      - run: AUTON_PACKAGE=auton python -m pip install '.[textual]'
+      - name: Test installed interface outside source checkout
+        run: |
+          cd "$(mktemp -d)"
+          python "$GITHUB_WORKSPACE/.github/scripts/check-test-collection.py" --runner unittest "$GITHUB_WORKSPACE/textual_tests"
+          python -m unittest discover -s "$GITHUB_WORKSPACE/textual_tests" -v
+      - name: Generate candidate captures from actual interface
+        run: PYTHONPATH=. python scripts/capture_textual.py --output "$RUNNER_TEMP/textual-captures"
+      - uses: actions/upload-artifact@v4
+        with:
+          name: textual-candidate-captures
+          path: ${{ runner.temp }}/textual-captures
+          if-no-files-found: error
+
   tests:
     runs-on: ubuntu-latest
     strategy:

AGENTS.md

@@ -119,3 +119,13 @@ Current project runner: `unittest` for `tests`. CI helper tests use
   with warnings treated as errors where supported. Review README/package text and
   the deployed manual for accidental maintainer content. Preserve private-project
   visibility and existing review/publication approval requirements.
+
+## Optional Textual presentation
+
+- Reuse `dwho.tui.textual` for shared layout, widgets and visual states. Keep
+  collection, permissions, rules and execution in application/client services.
+- Maintain the separate `textual_tests` unittest discovery root and optional CI
+  job. Base commands must work without Textual/Rich imports.
+- Generate candidate captures with `scripts/capture_textual.py`; keep synthetic
+  fixtures labelled and record source hashes. Publish shared website captures
+  only after the corresponding interface and source revision are accepted.

CHANGELOG

@@ -1,3 +1,11 @@
+auton (1.3.0) unstable; urgency=medium
+
+  * Add optional read-only Textual views backed by shared DWho components.
+  * Preserve existing default interface and service authorization boundaries.
+  * Include offline demo, separate UI tests and reproducible captures.
+
+ -- Codex <codex@openai.com>  Wed, 07 Oct 2026 02:00:00 +0200
+
 auton (1.2.1) unstable; urgency=medium
 
   * Hide unexpected job-service exception details in HTTP responses and logs.

MANIFEST.in

@@ -3,3 +3,6 @@ include requirements*.txt
 include bin/*
 recursive-include etc *
 recursive-include auton/web *.html *.css *.js *.svg
+
+recursive-include textual_tests *.py
+include scripts/capture_textual.py

README.md

@@ -654,7 +654,7 @@ Existing single-command and legacy failover invocations remain available.
 
 ### Operator TUI (client)
 
-The Unix client includes a ncurses interface with read-only monitoring and an
+The Unix client includes Textual and curses interfaces. See the [Textual guide](docs/textual.md) for the modern interface. The following keyboard reference describes curses, with read-only monitoring and an
 explicit execution preparation screen. It requires an interactive
 terminal with Python curses support and a daemon with the visibility API above.
 
@@ -750,7 +750,7 @@ from its pinned source on every website build and shared by GitHub and the web
 manual. [Capture provenance](https://auton.run/manual-captures/manifest.json)
 records the source commit, version and image checksums.
 
-Captures below come from real terminal sessions against disposable local daemons:
+Captures below show the real Textual interface with synthetic offline fixtures; no remote command is executed:
 
 [![Prepare targets and scenario groups](https://auton.run/manual-captures/tui-prepare.png)](https://auton.run/manual-captures/tui-prepare.png)
 [![Review the execution plan](https://auton.run/manual-captures/tui-confirm.png)](https://auton.run/manual-captures/tui-confirm.png)
@@ -781,9 +781,9 @@ without persistent central history or any change to execution failover.
 
 ### TUI screenshots
 
-These captures come from the real curses UI connected to two disposable local
-daemons. All job names and output are demonstration data. Click a capture to
-open the full-size image; the PNG files are stored in this repository.
+These captures show the Textual interface with labelled synthetic fixtures.
+All job names and output are demonstration data. Click a capture to open its
+full-size image; the website regenerates the shared PNG files from pinned sources.
 
 **Aggregated jobs**, including successful and failed commands:
 
@@ -1916,3 +1916,16 @@ coverage and compatibility.
 
 - **Users:** installation, configuration, operation and API usage in this README and the user guide.
 - **Contributors:** [architecture, tests and development](https://github.com/decryptus/auton/blob/master/CONTRIBUTING.md).
+
+
+## Textual terminal interface
+
+Install the optional extra and select the modern interface explicitly:
+
+```sh
+python -m pip install 'auton[textual]'
+auton --tui --tui-ui textual -c inventory.yml
+```
+
+See the [Textual guide](docs/textual.md) for navigation and supported operations.
+The existing curses interface remains available. Screenshots use synthetic data.

RELEASE

@@ -1 +1 @@
-1.2.1
+1.3.0

VERSION

@@ -1 +1 @@
-1.2.1
+1.3.0

auton_client/__init__.py

@@ -12,7 +12,7 @@
 from sonicprobe.libs import urisup
 
 SYSLOG_NAME = 'auton'
-__version__ = '1.2.1'
+__version__ = '1.3.0'
 DEFAULT_DELAY = 0.5
 DEFAULT_HTTP_TIMEOUT = 30
 

auton_client/preparation.py

@@ -15,108 +15,13 @@
 PAYLOAD_FIELDS = frozenset(('args', 'env'))
 
 
-class PreparationView:
-    def __init__(self, connections, groups, scenarios, scenario_groups, session, settings):
-        self.connections = dict(connections)
-        self.groups = {name: members for name, members in groups.items()
-                       if all(member in connections for member in members)}
-        self.scenarios = scenarios
-        self.scenario_groups = {name: members for name, members in scenario_groups.items()
-                                if all(member in scenarios for member in members)}
-        self.session, self.settings = session, settings
-        self.selected = [[] for _ in SECTIONS]
-        self.section = self.index = self.scroll = 0
-        self.search = ''
-        self.editing = None
-        self.input = '{"args": [], "env": {}}'
-        self.mode = 'select'
-        self.error = ''
-        self.preview = []
-        self.service = None
-        self.result = None
-        self.result_lines = None
-        self.endpoints = []
-        self.progress = {'revision': 0, 'rows': [], 'operation_id': None}
-        self.export_path = ''
-        self.import_path = ''
-        self.parameter_catalogue = {}
-        self.parameter_fields = []
-        self.parameter_values = []
-        self.parameter_index = 0
-        self.parameter_input = ''
+from auton_client.preparation_model import PreparationModel
+
+
+class PreparationView(PreparationModel):
 
-    def update(self, snapshots):
-        self.endpoints = sorted({item['name'] for data in snapshots.values()
-                                 for item in data.get('endpoints', [])})
-        catalogues = {}
-        for data in snapshots.values():
-            for item in data.get('endpoints', []):
-                catalogues.setdefault(item['name'], set()).add(json.dumps(item.get('parameters'), sort_keys=True))
-        self.parameter_catalogue = {name: json.loads(next(iter(values)))
-                                    for name, values in catalogues.items() if len(values) == 1}
-        if self.mode == 'running':
-            self.progress = self.session.progress.snapshot()
-        try:
-            result = self.session.poll()
-        except Exception:
-            self.error = 'Unable to collect operation result; do not replay. Inspect daemon jobs.'
-            self.mode = 'result'
-        else:
-            if result is not None:
-                self.result = result
-                self.result_lines = None
-                self.mode = 'result'
-                self.scroll = 0
 
-    def rows(self):
-        catalogues = (self.connections, self.groups, self.scenarios, self.scenario_groups, self.endpoints)
-        return [name for name in catalogues[self.section] if self.search.casefold() in name.casefold()]
 
-    def prepare(self):
-        targets = select_connections(self.selected[0], self.selected[1], self.connections, self.groups)
-        if not targets:
-            raise ValueError('Select at least one target or target group')
-        preview = ['Targets:']
-        for name, value in targets.items():
-            origins = target_origins(value)
-            preview.append('  %s = %s' % (name, origins[0]))
-            if len(origins) > 1:
-                preview.append('    Ordered failover (one destination, no broadcast):')
-                preview.extend('      %s' % origin for origin in origins[1:])
-        if self.selected[2] or self.selected[3]:
-            scenarios = select_scenarios(self.selected[2], self.selected[3], self.scenarios, self.scenario_groups)
-            service = ScenarioService(targets, scenarios, **self.settings)
-            for name, definition in scenarios.items():
-                preview.append('Scenario: ' + name)
-                for step in definition['steps']:
-                    preview.append('  %s -> %s' % (step['name'], step['endpoint']))
-                    preview.append('    args=' + json.dumps(step.get('args', [])))
-                    preview.append('    env=' + json.dumps(step.get('env', {})))
-        else:
-            if len(self.selected[4]) != 1:
-                raise ValueError('Select one endpoint or one or more scenarios')
-            try:
-                payload = json.loads(self.input)
-            except ValueError:
-                raise ValueError('Inputs must be valid JSON') from None
-            if (not isinstance(payload, dict) or not set(payload) <= PAYLOAD_FIELDS
-                    or not isinstance(payload.get('args', []), list)
-                    or any(not isinstance(value, str) or '\x00' in value for value in payload.get('args', []))
-                    or not isinstance(payload.get('env', {}), dict)
-                    or any(not isinstance(key, str) or not key or '=' in key or '\x00' in key
-                           or not isinstance(value, str) or '\x00' in value
-                           for key, value in payload.get('env', {}).items())):
-                raise ValueError('Inputs require args: string list and env: string mapping')
-            endpoint = self.selected[4][0]
-            service = OperationService(targets, endpoint, payload=payload, **self.settings)
-            preview.extend(['Endpoint: ' + endpoint, 'Inputs: ' + json.dumps(payload)])
-        preview.extend(['Parallel targets: %s; observation deadline: %ss' %
-                        (service.settings.parallel if isinstance(service, ScenarioService) else service.parallel,
-                         self.settings.get('timeout', 300)),
-                        'Server authentication and endpoint ACLs remain authoritative.',
-                        'No automatic replay, rollback or remote cancellation.'])
-        self.service, self.preview = service, preview
-        self.mode, self.scroll = 'confirm', 0
 
     def handle(self, key):
         if self.editing == 'parameter_input':

auton_client/preparation_model.py

@@ -0,0 +1,126 @@
+# -*- coding: utf-8 -*-
+# SPDX-License-Identifier: GPL-3.0-or-later
+"""Interface-neutral operation preparation; execution lives in services."""
+import json
+
+from auton_client.operations import OperationService
+from auton_client.export import export_result
+from auton_client.reconcile import ReconciliationService, load_report
+from auton_client.scenarios import ScenarioService, select_scenarios
+from auton_client.connections import select_connections, target_origins
+
+SECTIONS = ('Targets', 'Target groups', 'Scenarios', 'Scenario groups', 'Endpoints')
+MAX_INPUT = 4096
+PAYLOAD_FIELDS = frozenset(('args', 'env'))
+
+
+class PreparationModel:
+    def __init__(self, connections, groups, scenarios, scenario_groups, session, settings):
+        self.connections = dict(connections)
+        self.groups = {name: members for name, members in groups.items()
+                       if all(member in connections for member in members)}
+        self.scenarios = scenarios
+        self.scenario_groups = {name: members for name, members in scenario_groups.items()
+                                if all(member in scenarios for member in members)}
+        self.session, self.settings = session, settings
+        self.selected = [[] for _ in SECTIONS]
+        self.section = self.index = self.scroll = 0
+        self.search = ''
+        self.editing = None
+        self.input = '{"args": [], "env": {}}'
+        self.mode = 'select'
+        self.error = ''
+        self.preview = []
+        self.service = None
+        self.result = None
+        self.result_lines = None
+        self.endpoints = []
+        self.progress = {'revision': 0, 'rows': [], 'operation_id': None}
+        self.export_path = ''
+        self.import_path = ''
+        self.parameter_catalogue = {}
+        self.parameter_fields = []
+        self.parameter_values = []
+        self.parameter_index = 0
+        self.parameter_input = ''
+
+
+    def update(self, snapshots):
+        self.endpoints = sorted({item['name'] for data in snapshots.values()
+                                 for item in data.get('endpoints', [])})
+        catalogues = {}
+        for data in snapshots.values():
+            for item in data.get('endpoints', []):
+                catalogues.setdefault(item['name'], set()).add(json.dumps(item.get('parameters'), sort_keys=True))
+        self.parameter_catalogue = {name: json.loads(next(iter(values)))
+                                    for name, values in catalogues.items() if len(values) == 1}
+        self.poll_result()
+
+    def poll_result(self):
+        if self.mode == 'running':
+            self.progress = self.session.progress.snapshot()
+        try:
+            result = self.session.poll()
+        except Exception:
+            self.error = 'Unable to collect operation result; do not replay. Inspect daemon jobs.'
+            self.mode = 'result'
+        else:
+            if result is not None:
+                self.result = result
+                self.result_lines = None
+                self.mode = 'result'
+                self.scroll = 0
+
+
+    def rows(self):
+        catalogues = (self.connections, self.groups, self.scenarios, self.scenario_groups, self.endpoints)
+        return [name for name in catalogues[self.section] if self.search.casefold() in name.casefold()]
+
+
+    def prepare(self):
+        targets = select_connections(self.selected[0], self.selected[1], self.connections, self.groups)
+        if not targets:
+            raise ValueError('Select at least one target or target group')
+        preview = ['Targets:']
+        for name, value in targets.items():
+            origins = target_origins(value)
+            preview.append('  %s = %s' % (name, origins[0]))
+            if len(origins) > 1:
+                preview.append('    Ordered failover (one destination, no broadcast):')
+                preview.extend('      %s' % origin for origin in origins[1:])
+        if self.selected[2] or self.selected[3]:
+            scenarios = select_scenarios(self.selected[2], self.selected[3], self.scenarios, self.scenario_groups)
+            service = ScenarioService(targets, scenarios, **self.settings)
+            for name, definition in scenarios.items():
+                preview.append('Scenario: ' + name)
+                for step in definition['steps']:
+                    preview.append('  %s -> %s' % (step['name'], step['endpoint']))
+                    preview.append('    args=' + json.dumps(step.get('args', [])))
+                    preview.append('    env=' + json.dumps(step.get('env', {})))
+        else:
+            if len(self.selected[4]) != 1:
+                raise ValueError('Select one endpoint or one or more scenarios')
+            try:
+                payload = json.loads(self.input)
+            except ValueError:
+                raise ValueError('Inputs must be valid JSON') from None
+            if (not isinstance(payload, dict) or not set(payload) <= PAYLOAD_FIELDS
+                    or not isinstance(payload.get('args', []), list)
+                    or any(not isinstance(value, str) or '\x00' in value for value in payload.get('args', []))
+                    or not isinstance(payload.get('env', {}), dict)
+                    or any(not isinstance(key, str) or not key or '=' in key or '\x00' in key
+                           or not isinstance(value, str) or '\x00' in value
+                           for key, value in payload.get('env', {}).items())):
+                raise ValueError('Inputs require args: string list and env: string mapping')
+            endpoint = self.selected[4][0]
+            service = OperationService(targets, endpoint, payload=payload, **self.settings)
+            preview.extend(['Endpoint: ' + endpoint, 'Inputs: ' + json.dumps(payload)])
+        preview.extend(['Parallel targets: %s; observation deadline: %ss' %
+                        (service.settings.parallel if isinstance(service, ScenarioService) else service.parallel,
+                         self.settings.get('timeout', 300)),
+                        'Server authentication and endpoint ACLs remain authoritative.',
+                        'No automatic replay, rollback or remote cancellation.'])
+        self.service, self.preview = service, preview
+        self.mode, self.scroll = 'confirm', 0
+
+