chore: release v0.17.6

- Add dependency-aware wave execution, prompt input, and output persistence to aur spawn
- Fix truncation of long goals content before embedding (2048 char limit)
- Replace HTML comment agent metadata with visible markdown sub-bullets
- Remove 98 unused imports across test and source files
- Fix pre-commit config: exclude fixture broken.py, add E731 ignore
- Fix D212 docstring style in scripts

Author: hamr0

.pre-commit-config.yaml

@@ -35,7 +35,7 @@ repos:
 
       - id: debug-statements
         name: Check for debug statements (pdb, ipdb)
-        exclude: '^tests/'
+        exclude: '^(tests/|.*fixtures/sample_python_files/broken\.py$)'
 
       - id: mixed-line-ending
         name: Check for mixed line endings
@@ -48,7 +48,7 @@ repos:
       - id: black
         name: Format Python code with Black
         args: ['--line-length=100']
-        exclude: 'tests/fixtures/.*broken\.py$'
+        exclude: '.*fixtures/.*broken\.py$'
 
   # Python import sorting
   - repo: https://github.com/pycqa/isort
@@ -66,8 +66,8 @@ repos:
         name: Lint Python code with flake8
         args: [
           '--max-line-length=100',
-          '--extend-ignore=E203,W503,E501,E722,E721,F841,E402,F541',  # Black compatibility + test/script leniency
-          '--exclude=.git,__pycache__,build,dist,*.egg-info,tests/fixtures/sample_python_files/broken.py'
+          '--extend-ignore=E203,W503,E501,E722,E721,F841,E402,F541,E731',  # Black compatibility + test/script leniency
+          '--exclude=.git,__pycache__,build,dist,*.egg-info,**/fixtures/sample_python_files/broken.py'
         ]
 
   # Python type checking (optional, can be slow)
@@ -125,7 +125,7 @@ repos:
       - id: pydocstyle
         name: Check docstring style
         args: ['--convention=google', '--add-ignore=D100,D101,D102,D103,D104,D105,D106,D107,D202,D205,D209,D300,D301,D402,D403,D415']
-        exclude: '^(tests/|packages/.*/tests/|docs/archive/|.*_test\.py$|test_.*\.py$|compare_batch_sizes\.py|profile_.*\.py|benchmark_.*\.py|saas_financial_model\.py)'
+        exclude: '^(tests/|packages/.*/tests/|packages/.*/fixtures/|docs/archive/|.*_test\.py$|test_.*\.py$|compare_batch_sizes\.py|profile_.*\.py|benchmark_.*\.py|saas_financial_model\.py)'
 
 # Configuration
 default_language_version:

CHANGELOG.md

@@ -7,6 +7,35 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.17.6] - 2026-02-14
+
+### Added
+
+- **`aur spawn` dependency-aware wave execution**
+  - `depends_on` field in `ParsedTask` with `- Depends: 1.0, 2.0` sub-bullet parsing
+  - Topological sort (Kahn's algorithm) producing parallel-safe waves
+  - Wave-based execution: sequential across waves, parallel within each wave
+  - Completed task outputs forwarded as context into dependent tasks
+  - Dry-run shows wave breakdown
+- **`aur spawn` prompt-to-tasks decomposition**
+  - `decompose_prompt_to_tasks_md()` via LLM for natural language → tasks.md
+  - Auto-detection of input type: file path vs prompt text
+  - Discovers available agents via `AgentScanner`/`AgentParser`
+- **`aur spawn` output persistence**
+  - `SpawnRunStore` class with `.aurora/spawn/runs/<timestamp>/` structure
+  - Stores `tasks.md`, `results/task-{id}.json`, `summary.json`, `meta.json`
+  - Re-run detection via SHA-256 hash; skip-completed-tasks on re-runs
+- `FEATURE_BACKLOG.md` tracking parked features and future work
+
+### Fixed
+
+- Truncate long goals content before embedding to avoid 2048 char limit (full content kept for BM25/FTS5)
+
+### Changed
+
+- Replace HTML comment agent metadata with visible markdown sub-bullets in task output
+- Code formatting cleanup across 170+ files (black/isort)
+
 ## [0.17.2] - 2026-02-14
 
 ### Fixed

packages/cli/src/aurora_cli/commands/spawn.py

@@ -41,7 +41,6 @@
 from implement.persistence import SpawnRunStore
 from implement.topo_sort import topological_sort_tasks
 
-
 console = Console()
 logger = logging.getLogger(__name__)
 
@@ -204,7 +203,9 @@ def spawn_command(
 
         try:
             if has_deps:
-                console.print(f"[cyan]Executing {len(pending_tasks)} tasks in dependency waves...[/]")
+                console.print(
+                    f"[cyan]Executing {len(pending_tasks)} tasks in dependency waves...[/]"
+                )
                 result = asyncio.run(
                     _execute_waves(
                         pending_tasks,
@@ -345,7 +346,9 @@ def _display_dry_run(tasks: list[ParsedTask], has_deps: bool) -> None:
                 for task in wave:
                     status = "[x]" if task.completed else "[ ]"
                     deps = f" (depends: {', '.join(task.depends_on)})" if task.depends_on else ""
-                    console.print(f"    {status} {task.id}. {task.description} (agent: {task.agent}){deps}")
+                    console.print(
+                        f"    {status} {task.id}. {task.description} (agent: {task.agent}){deps}"
+                    )
         except ValueError as e:
             console.print(f"  [red]Dependency error: {e}[/]")
     else:
@@ -674,4 +677,6 @@ def execute_tasks_parallel(tasks: list[ParsedTask]) -> dict[str, int]:
         Execution summary with total, completed, failed counts
 
     """
-    return asyncio.run(_execute_parallel(tasks, verbose=False, store=SpawnRunStore(), run_dir=Path(".")))
+    return asyncio.run(
+        _execute_parallel(tasks, verbose=False, store=SpawnRunStore(), run_dir=Path("."))
+    )

packages/cli/src/aurora_cli/main.py

@@ -30,7 +30,7 @@
 console = Console()
 logger = logging.getLogger(__name__)
 
-AURORA_VERSION = "0.17.2"
+AURORA_VERSION = "0.17.6"
 
 
 def _version_callback(ctx: click.Context, param: click.Parameter, value: bool) -> None:

packages/cli/tests/integration/test_escalation_health.py

@@ -5,9 +5,6 @@
 No LLM calls — escalation uses keyword-only mode, health checks use mocked filesystem.
 """
 
-import os
-import shutil
-
 import pytest
 
 from aurora_cli.escalation import AutoEscalationHandler, EscalationConfig, EscalationResult

packages/cli/tests/test_config.py

@@ -7,7 +7,6 @@
 """
 
 import json
-import os
 
 import pytest
 

packages/context-code/src/aurora_context_code/semantic/hybrid_retriever.py

@@ -26,7 +26,6 @@
 import time
 from collections import OrderedDict
 from dataclasses import dataclass
-from pathlib import Path
 from typing import Any
 
 import numpy as np

packages/context-code/tests/integration/test_fts5_retrieval.py

@@ -4,7 +4,6 @@
 """
 
 import numpy as np
-import pytest
 
 from aurora_context_code.semantic.hybrid_retriever import HybridConfig, HybridRetriever
 from aurora_core.store.sqlite import SQLiteStore
@@ -30,8 +29,6 @@ def _make_code_chunk(chunk_id, name, signature, docstring, file_path):
 class MockActivationEngine:
     """Minimal activation engine for testing."""
 
-    pass
-
 
 class MockEmbeddingProvider:
     """Embedding provider that returns deterministic embeddings."""

packages/context-code/tests/integration/test_retriever_fallbacks.py

@@ -5,8 +5,6 @@
 when embeddings are unavailable.
 """
 
-import math
-
 import numpy as np
 import pytest
 

packages/core/src/aurora_core/chunk_types.py

@@ -11,7 +11,6 @@
 """
 
 from pathlib import Path
-from typing import Union
 
 # Extension-based type mapping
 EXTENSION_TYPE_MAP: dict[str, str] = {