QA-Improver

.claude/agents/qa-improver.md

@@ -0,0 +1,85 @@
+---
+name: qa-improver
+description: Automatically fix code quality issues identified by a QA assessment. Resolves lint violations, type errors, missing exports, and other mechanical improvements without changing behavior. Use after a quality-scorer run to action its recommendations.
+tools: Bash Read Write Edit Grep Glob
+---
+
+You are a code quality improvement agent for the Impakt project. You receive a QA report and systematically fix the mechanical issues it identifies — without changing any runtime behavior.
+
+## Inputs
+
+Read the most recent `docs/QA-*.md` report (excluding QA-TEMPLATE.md and QA-INSTRUCTIONS.md). Extract the "Recommended Actions" table.
+
+## Workflow
+
+### 1. Assess what can be auto-fixed
+
+Categorize each recommended action:
+
+- **Auto-fixable**: lint auto-fix, import sorting, unused imports/vars
+- **Mechanical**: adding type annotations, `__all__` exports, duplicate dict keys, line-length fixes
+- **Requires judgment**: refactoring complex files, adding test coverage, security changes
+
+Only perform auto-fixable and mechanical fixes. Skip anything that requires judgment or behavioral changes.
+
+### 2. Execute fixes in order of safety
+
+Run fixes from safest to most involved:
+
+**a) Lint auto-fix (safest)**
+```bash
+uv run python -m ruff check --fix src/
+```
+
+**b) Remaining lint violations**
+Read the ruff output. For each remaining violation:
+- F601 (duplicate dict keys): read the file, determine which entry to keep, remove the duplicate
+- F841 (unused variables): remove or prefix with `_`
+- E501 (line too long): break the line naturally
+- F541 (empty f-string): convert to regular string
+
+**c) Type annotation fixes**
+Run `uv run python -m mypy src/impakt --ignore-missing-imports` and fix `[type-arg]` errors by adding proper generic parameters (e.g., `dict` -> `dict[str, Any]`, `list` -> `list[str]`). Read context around each error to determine the correct type.
+
+**d) Missing `__all__` exports**
+For any module `__init__.py` that lacks `__all__`:
+- List the public classes and functions in the module's files
+- Add imports and `__all__` following the pattern in `src/impakt/channel/__init__.py`
+
+**e) Coverage config (if not already present)**
+Check if `addopts` in `[tool.pytest.ini_options]` includes `--cov`. If not, add it.
+
+### 3. Verify after each category
+
+After each category of fixes, run:
+```bash
+uv run python -m ruff check src/
+uv run python -m pytest --tb=short -q
+```
+
+If tests fail, revert the last change and move on. Never leave the codebase in a broken state.
+
+### 4. Final verification
+
+Run all three quality tools:
+```bash
+uv run python -m ruff check src/
+uv run python -m mypy src/impakt --ignore-missing-imports
+uv run python -m pytest --tb=short -q
+```
+
+### 5. Report results
+
+Return a summary:
+- Which actions were completed
+- Which were skipped and why
+- Before/after counts for each tool (lint violations, mypy errors, test results)
+- Any actions that still require human judgment
+
+## Rules
+
+- **Never change runtime behavior.** Only annotations, imports, formatting, and dead code removal.
+- **Never modify test files.** Only source code under `src/`.
+- **Verify after every category.** If tests break, revert immediately.
+- **Be conservative with type annotations.** Use `Any` when the actual type is genuinely dynamic. Don't guess complex types — leave them for human review.
+- **Don't touch security-sensitive code** (eval, exec, subprocess) unless the QA report explicitly flags an unsafe pattern AND the fix is mechanical.

pyproject.toml

@@ -62,6 +62,7 @@ packages = ["src/impakt"]
 [tool.pytest.ini_options]
 testpaths = ["tests"]
 pythonpath = ["src"]
+addopts = "--cov=impakt --cov-report=term-missing:skip-covered --cov-fail-under=60"
 filterwarnings = [
     "ignore::pytest.PytestCollectionWarning",
 ]

scripts/qa-improve.sh

@@ -0,0 +1,136 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+# ──────────────────────────────────────────────────────────────
+# qa-improve.sh — Score the codebase, then auto-fix what's found
+#
+# Runs two agents in sequence:
+#   1. quality-scorer  → produces a QA report in docs/
+#   2. qa-improver     → reads the report and fixes mechanical issues
+#   3. quality-scorer  → re-scores to measure improvement
+#
+# Usage:
+#   ./scripts/qa-improve.sh                # default: sonnet
+#   ./scripts/qa-improve.sh --model opus   # use opus for deeper analysis
+#   ./scripts/qa-improve.sh --score-only   # skip the improvement step
+#   ./scripts/qa-improve.sh --fix-only     # skip initial scoring (use latest report)
+# ──────────────────────────────────────────────────────────────
+
+PROJECT_ROOT="$(cd "$(dirname "$0")/.." && pwd)"
+MODEL="sonnet"
+BUDGET="1.50"
+SKIP_SCORE=false
+SKIP_FIX=false
+
+while [[ $# -gt 0 ]]; do
+  case "$1" in
+    --model)      MODEL="$2"; shift 2 ;;
+    --budget)     BUDGET="$2"; shift 2 ;;
+    --score-only) SKIP_FIX=true; shift ;;
+    --fix-only)   SKIP_SCORE=true; shift ;;
+    --help|-h)
+      echo "Usage: $0 [--model sonnet|opus|haiku] [--budget USD] [--score-only] [--fix-only]"
+      echo ""
+      echo "Score the codebase, auto-fix issues, then re-score to measure improvement."
+      echo ""
+      echo "Options:"
+      echo "  --model MODEL    Claude model (default: sonnet)"
+      echo "  --budget USD     Max spend per agent run (default: 1.50)"
+      echo "  --score-only     Run scoring only, skip auto-fix"
+      echo "  --fix-only       Skip initial score, fix based on latest report"
+      exit 0
+      ;;
+    *) echo "Unknown option: $1"; exit 1 ;;
+  esac
+done
+
+cd "$PROJECT_ROOT"
+
+# Verify prerequisites
+for cmd in claude uv; do
+  if ! command -v "$cmd" &>/dev/null; then
+    echo "Error: $cmd not found." >&2
+    exit 1
+  fi
+done
+
+if [[ ! -f "docs/QA-INSTRUCTIONS.md" ]]; then
+  echo "Error: docs/QA-INSTRUCTIONS.md not found." >&2
+  exit 1
+fi
+
+uv sync --dev --quiet
+
+SCORE_PROMPT="Run a full codebase quality assessment. \
+Read docs/QA-INSTRUCTIONS.md for the methodology and rubrics. \
+Read docs/QA-TEMPLATE.md for the report structure. \
+Check docs/ for previous QA-*.md reports and compute deltas if any exist. \
+Collect all raw metrics by running every command in Step 1. \
+Score each dimension using the Step 2 rubrics. \
+Compute the composite score using the Step 3 formula. \
+Write the completed report to docs/QA-<datetime>.md. \
+Print the composite score, grade, per-dimension scores, and top 3 actions."
+
+FIX_PROMPT="Read the most recent QA report in docs/ (the QA-*.md file with the latest date, \
+excluding QA-TEMPLATE.md and QA-INSTRUCTIONS.md). \
+Extract the Recommended Actions table. \
+Execute all auto-fixable and mechanical fixes: \
+1) Run uv run ruff check --fix src/ for lint auto-fixes. \
+2) Fix remaining lint violations (duplicate keys, unused vars, long lines). \
+3) Fix mypy type-arg errors by adding proper generic parameters. \
+4) Add __all__ exports to any modules missing them. \
+5) Verify after each category with ruff check and pytest. \
+Never change runtime behavior. Never modify test files. \
+Report what was fixed and what was skipped."
+
+# ── Phase 1: Score ──
+if [[ "$SKIP_SCORE" == false ]]; then
+  echo "Phase 1: Scoring codebase (model: $MODEL)..."
+  echo "────────────────────────────────────────────"
+
+  claude -p \
+    --agent quality-scorer \
+    --model "$MODEL" \
+    --max-budget-usd "$BUDGET" \
+    --allowedTools "Bash Read Write Grep Glob" \
+    --output-format text \
+    "$SCORE_PROMPT"
+
+  REPORT=$(ls -t docs/QA-2*.md 2>/dev/null | head -1)
+  echo ""
+  echo "Report: ${REPORT:-none found}"
+  echo ""
+fi
+
+# ── Phase 2: Fix ──
+if [[ "$SKIP_FIX" == false ]]; then
+  echo "Phase 2: Auto-fixing issues (model: $MODEL)..."
+  echo "────────────────────────────────────────────"
+
+  claude -p \
+    --agent qa-improver \
+    --model "$MODEL" \
+    --max-budget-usd "$BUDGET" \
+    --allowedTools "Bash Read Write Edit Grep Glob" \
+    --output-format text \
+    "$FIX_PROMPT"
+
+  echo ""
+
+  # ── Phase 3: Re-score ──
+  echo "Phase 3: Re-scoring after fixes (model: $MODEL)..."
+  echo "────────────────────────────────────────────"
+
+  claude -p \
+    --agent quality-scorer \
+    --model "$MODEL" \
+    --max-budget-usd "$BUDGET" \
+    --allowedTools "Bash Read Write Grep Glob" \
+    --output-format text \
+    "$SCORE_PROMPT"
+
+  FINAL_REPORT=$(ls -t docs/QA-2*.md 2>/dev/null | head -1)
+  echo ""
+  echo "────────────────────────────────────────────"
+  echo "Final report: ${FINAL_REPORT:-none found}"
+fi

src/impakt/channel/lookup.py

@@ -60,10 +60,9 @@ MAIN_LOCATIONS: dict[str, str] = {
     "CLAV": "Clavicle",
     "RIBS": "Ribs",
     "STRN": "Sternum",
-    "SHLR": "Shoulder",
+    "SHLR": "Shoulder Right",
     "SHLD": "Shoulder",
     "SHLL": "Shoulder Left",
-    "SHLR": "Shoulder Right",
     # Spine
     "SPIN": "Spine",
     "LUSP": "Lumbar Spine",
@@ -74,7 +73,6 @@ MAIN_LOCATIONS: dict[str, str] = {
     "SACR": "Sacrum",
     "PUBC": "Pubic Symphysis",
     # Upper extremities
-    "SHLD": "Shoulder",
     "UPRA": "Upper Arm",
     "ELBO": "Elbow",
     "FORA": "Forearm",
@@ -122,7 +120,7 @@ MAIN_LOCATIONS: dict[str, str] = {
     # Wheel
     "WHEL": "Wheel",
     # Structural — additional
-    "FORA": "Floor Rail",
+    "FRAL": "Floor Rail",
     "FBAR": "Barrier Face",
     "KNSL": "Knee Slider",
     # Simulation / other
@@ -181,11 +179,6 @@ FINE_LOCATIONS: dict[str, str] = {
     "URXX": "Upper Right",
     "LLXX": "Lower Left",
     "LRXX": "Lower Right",
-    # Tibia positions (real MME data uses these)
-    "LEUP": "Left Upper",
-    "RIUP": "Right Upper",
-    "LELO": "Left Lower",
-    "RILO": "Right Lower",
     # Fine location with number suffix
     "0001": "Secondary",
     "0100": "Lower Position",

src/impakt/channel/model.py

@@ -8,17 +8,17 @@ from __future__ import annotations
 
 import fnmatch
 from collections import defaultdict
+from collections.abc import Iterator
 from dataclasses import dataclass, field
 from datetime import date
 from pathlib import Path
-from typing import Any, Iterator
+from typing import Any
 
 import numpy as np
 from numpy.typing import NDArray
 
 from impakt.channel.code import ChannelCode
 
-
 # ---------------------------------------------------------------------------
 # Metadata models
 # ---------------------------------------------------------------------------

src/impakt/criteria/chest.py

@@ -6,14 +6,11 @@ Viscous Criterion (VC): V(t) * C(t) where V = deflection velocity, C = compressi
 
 from __future__ import annotations
 
-from typing import Any
-
 import numpy as np
 
 from impakt.channel.model import Channel, DummyInfo
 from impakt.criteria.base import CriterionResult
 
-
 # Initial chest depth by dummy type (mm)
 CHEST_DEPTH: dict[str, float] = {
     "H3-50M": 229.0,

src/impakt/criteria/clip3ms.py

@@ -7,8 +7,6 @@ where the total time the signal exceeds that level equals at least 3 ms.
 
 from __future__ import annotations
 
-from typing import Any
-
 import numpy as np
 
 from impakt.channel.model import Channel, ChannelGroup, DummyInfo

src/impakt/criteria/femur.py

@@ -6,8 +6,6 @@ Evaluated separately for left and right femur.
 
 from __future__ import annotations
 
-from typing import Any
-
 import numpy as np
 
 from impakt.channel.model import Channel, DummyInfo

src/impakt/criteria/nij.py

@@ -14,7 +14,6 @@ The reported Nij is the maximum across all four modes and all time steps.
 from __future__ import annotations
 
 from dataclasses import dataclass
-from typing import Any
 
 import numpy as np
 

src/impakt/criteria/tibia.py

@@ -12,7 +12,6 @@ where:
 from __future__ import annotations
 
 from dataclasses import dataclass
-from typing import Any
 
 import numpy as np
 

src/impakt/io/__init__.py

@@ -1 +1,17 @@
 """Data I/O readers for crash test formats."""
+
+from impakt.io.csv import CSVReader
+from impakt.io.mme import MMEReader
+from impakt.io.reader import ReaderProtocol, ReaderRegistry, get_registry, read, register_reader
+from impakt.io.tdms import TDMSReader
+
+__all__ = [
+    "CSVReader",
+    "MMEReader",
+    "ReaderProtocol",
+    "ReaderRegistry",
+    "TDMSReader",
+    "get_registry",
+    "read",
+    "register_reader",
+]

src/impakt/io/mme.py

@@ -89,7 +89,6 @@ UNIT_MAP: dict[str, str] = {
     "Nm": "N·m",
     "N.m": "N·m",
     "N*m": "N·m",
-    "Nm": "N·m",
     "mm": "mm",
     "m": "m",
     "cm": "cm",
@@ -626,7 +625,7 @@ class MMEReader:
         unit = _normalize_unit(header.get("unit", ""))
         dt = _parse_float(header.get("sampling interval", "0"))
         t_first = _parse_float(header.get("time of first sample", "0"))
-        num_samples_declared = _parse_int(header.get("number of samples", "0"))
+        _parse_int(header.get("number of samples", "0"))  # validated by data length
         cfc_str = header.get("channel frequency class", "")
         cfc_class: int | None = None
         if cfc_str and cfc_str.upper() != "NOVALUE":

src/impakt/plot/cursor.py

@@ -6,8 +6,6 @@ across all plotted channels.
 
 from __future__ import annotations
 
-from typing import Any
-
 import numpy as np
 import pandas as pd
 

src/impakt/plot/engine.py

@@ -11,11 +11,10 @@ from __future__ import annotations
 
 from typing import Any
 
-import numpy as np
 import plotly.graph_objects as go
 
 from impakt.channel.model import Channel
-from impakt.plot.spec import ChannelRef, Corridor, CursorValues, PlotSpec, PlotStyle
+from impakt.plot.spec import Corridor, CursorValues, PlotSpec
 
 # Default color palette (colorblind-friendly)
 DEFAULT_COLORS = [

src/impakt/plot/export.py

@@ -3,7 +3,6 @@
 from __future__ import annotations
 
 from pathlib import Path
-from typing import Any
 
 from impakt.plot.engine import PlotEngine
 from impakt.plot.spec import PlotSpec

src/impakt/plugin/__init__.py

@@ -1 +1,19 @@
 """Plugin system for extensibility."""
+
+from impakt.plugin.registry import (
+    ImpaktPlugin,
+    PluginRegistry,
+    discover_all,
+    discover_directory,
+    discover_entry_points,
+    get_plugin_registry,
+)
+
+__all__ = [
+    "ImpaktPlugin",
+    "PluginRegistry",
+    "discover_all",
+    "discover_directory",
+    "discover_entry_points",
+    "get_plugin_registry",
+]

src/impakt/protocol/euro_ncap.py

@@ -10,7 +10,6 @@ Threshold values are versioned — this module supports multiple protocol years.
 from __future__ import annotations
 
 from pathlib import Path
-from typing import Any
 
 from impakt.criteria.base import CriterionResult
 from impakt.protocol.base import BodyRegionScore, Color, ProtocolResult

src/impakt/protocol/iihs.py

@@ -7,7 +7,6 @@ The overall rating is determined by the worst sub-rating.
 from __future__ import annotations
 
 from pathlib import Path
-from typing import Any
 
 from impakt.criteria.base import CriterionResult
 from impakt.protocol.base import BodyRegionScore, ProtocolResult, Rating

src/impakt/protocol/us_ncap.py

@@ -8,7 +8,6 @@ injury probability, which are then combined.
 from __future__ import annotations
 
 import math
-from typing import Any
 
 from impakt.criteria.base import CriterionResult
 from impakt.protocol.base import BodyRegionScore, ProtocolResult, Rating

src/impakt/report/__init__.py

@@ -1 +1,13 @@
 """PDF and report generation."""
+
+from impakt.report.engine import (
+    generate_injury_summary,
+    generate_plot_sheet,
+    generate_protocol_report,
+)
+
+__all__ = [
+    "generate_injury_summary",
+    "generate_plot_sheet",
+    "generate_protocol_report",
+]

src/impakt/report/engine.py

@@ -155,7 +155,8 @@ def _fallback_protocol_html(
         test_info = f"""
         <div class="test-info">
             <p><strong>Test:</strong> {metadata.test_number}</p>
-            <p><strong>Vehicle:</strong> {metadata.vehicle.year} {metadata.vehicle.make} {metadata.vehicle.model}</p>
+            <p><strong>Vehicle:</strong> {metadata.vehicle.year} {metadata.vehicle.make} \
+{metadata.vehicle.model}</p>
             <p><strong>Dummy:</strong> {metadata.dummy.dummy_type} ({metadata.dummy.position})</p>
         </div>
         """
@@ -178,7 +179,11 @@ def _fallback_protocol_html(
             color_badge = f'<span class="badge" style="background:{bg}">{rs.rating.value}</span>'
 
         points_str = f"{rs.points:.1f}/{rs.max_points:.1f}" if rs.max_points > 0 else ""
-        rows += f"<tr><td>{rs.region}</td><td>{rs.criterion}</td><td>{rs.value:.2f} {rs.unit}</td><td>{color_badge}</td><td>{points_str}</td></tr>"
+        rows += (
+            f"<tr><td>{rs.region}</td><td>{rs.criterion}</td>"
+            f"<td>{rs.value:.2f} {rs.unit}</td>"
+            f"<td>{color_badge}</td><td>{points_str}</td></tr>"
+        )
 
     stars_display = ""
     if result.stars is not None:
@@ -207,10 +212,12 @@ def _fallback_protocol_html(
         <div class="summary">
             <p><span class="stars">{stars_display}</span></p>
             <p><strong>Overall:</strong> {result.overall_rating}</p>
-            <p><strong>Score:</strong> {result.total_points:.1f}/{result.max_points:.1f} ({result.percentage:.0f}%)</p>
+            <p><strong>Score:</strong> \
+{result.total_points:.1f}/{result.max_points:.1f} ({result.percentage:.0f}%)</p>
         </div>
         <table>
-            <tr><th>Body Region</th><th>Criterion</th><th>Value</th><th>Rating</th><th>Points</th></tr>
+            <tr><th>Body Region</th><th>Criterion</th><th>Value</th>\
+<th>Rating</th><th>Points</th></tr>
             {rows}
         </table>
     </body>

src/impakt/script/__init__.py

@@ -1 +1,12 @@
 """Scripting API and CLI."""
+
+from impakt.script.api import ChannelHandle, Session, Template, TransformProxy
+from impakt.script.cli import main
+
+__all__ = [
+    "ChannelHandle",
+    "Session",
+    "Template",
+    "TransformProxy",
+    "main",
+]

src/impakt/script/api.py

@@ -32,12 +32,12 @@ import numpy as np
 from impakt.channel.model import Channel, ChannelGroup, TestData, TestMetadata
 from impakt.criteria.base import CriterionResult
 from impakt.io.mme import MMEReader
-from impakt.io.reader import ReaderRegistry, get_registry, register_reader
-from impakt.plot.engine import PlotEngine, cursor_values
-from impakt.plot.spec import ChannelRef, CursorValues, PlotSpec, PlotStyle
+from impakt.io.reader import get_registry, register_reader
+from impakt.plot.engine import PlotEngine
+from impakt.plot.spec import ChannelRef, PlotSpec, PlotStyle
 from impakt.protocol.base import ProtocolResult
 from impakt.template.library import TemplateLibrary
-from impakt.template.model import SessionState, TemplateSpec
+from impakt.template.model import TemplateSpec
 from impakt.template.session import SessionManager
 from impakt.transform.cfc import CFCFilter
 

src/impakt/script/cli.py

@@ -3,8 +3,6 @@
 from __future__ import annotations
 
 import argparse
-import sys
-from pathlib import Path
 
 
 def main(argv: list[str] | None = None) -> None:
@@ -121,7 +119,6 @@ def _cmd_channels(args: argparse.Namespace) -> None:
 
 def _cmd_evaluate(args: argparse.Namespace) -> None:
     from impakt.script.api import Session
-    from impakt.criteria import hic, clip_3ms, nij, chest_deflection, femur_load
 
     session = Session.open(args.path)
 

src/impakt/template/__init__.py

@@ -1 +1,13 @@
 """Template and session management."""
+
+from impakt.template.library import TemplateLibrary
+from impakt.template.model import PlotDefinition, SessionState, TemplateSpec
+from impakt.template.session import SessionManager
+
+__all__ = [
+    "PlotDefinition",
+    "SessionManager",
+    "SessionState",
+    "TemplateLibrary",
+    "TemplateSpec",
+]

src/impakt/template/library.py

@@ -6,8 +6,8 @@ Manages the global template library at ``~/.impakt/templates/``.
 from __future__ import annotations
 
 import logging
+from collections.abc import Iterator
 from pathlib import Path
-from typing import Iterator
 
 from impakt.template.model import TemplateSpec
 

src/impakt/template/model.py

@@ -6,7 +6,6 @@ Sessions bind templates to specific test data and track user overrides.
 
 from __future__ import annotations
 
-import copy
 from dataclasses import dataclass, field
 from datetime import datetime
 from pathlib import Path
@@ -94,7 +93,6 @@ class TemplateSpec:
         for plot_data in data.get("plots", []):
             channel_patterns = []
             transforms = []
-            corridors = []
 
             for ch in plot_data.get("channels", []):
                 if isinstance(ch, str):

src/impakt/transform/align.py

@@ -76,7 +76,7 @@ class XAlign:
             # Threshold never crossed — no shift
             return channel.with_data(
                 data=channel.data,
-                transform_note=f"X-align threshold (no crossing found)",
+                transform_note="X-align threshold (no crossing found)",
             )
 
         crossing_time = float(channel.time[indices[0]])

src/impakt/transform/base.py

@@ -7,7 +7,7 @@ TransformChain composes multiple transforms into a reproducible pipeline.
 
 from __future__ import annotations
 
-from dataclasses import dataclass, field
+from dataclasses import dataclass
 from typing import Any, Protocol, runtime_checkable
 
 from impakt.channel.model import Channel

src/impakt/transform/cfc.py

@@ -15,13 +15,11 @@ from __future__ import annotations
 from dataclasses import dataclass
 from typing import Any
 
-import numpy as np
 from scipy.signal import butter, filtfilt
 
 from impakt.channel.lookup import CFC_CLASSES, CFC_MINIMUM_SAMPLE_RATES
 from impakt.channel.model import Channel
 
-
 VALID_CFC_CLASSES = frozenset(CFC_CLASSES.keys())
 
 

src/impakt/transform/math_expr.py

@@ -7,7 +7,6 @@ with numpy functions.
 
 from __future__ import annotations
 
-import re
 from dataclasses import dataclass
 from typing import Any
 
@@ -16,7 +15,6 @@ import numpy as np
 from impakt.channel.code import ChannelCode
 from impakt.channel.model import Channel
 
-
 # Allowed numpy functions in expressions
 _SAFE_NUMPY = {
     "abs": np.abs,

src/impakt/web/app.py

@@ -6,9 +6,6 @@ The AppState holds Session objects server-side; Dash stores hold lightweight UI
 
 from __future__ import annotations
 
-from pathlib import Path
-from typing import TYPE_CHECKING
-
 import dash
 import dash_bootstrap_components as dbc
 

src/impakt/web/callbacks/channel_callbacks.py

@@ -13,8 +13,7 @@ from __future__ import annotations
 from typing import Any
 
 import dash
-from dash import ALL, Input, Output, State, html, no_update
-from dash.exceptions import PreventUpdate
+from dash import ALL, Input, Output, State
 
 from impakt.plot.engine import DEFAULT_COLORS
 from impakt.web.components.channel_grid import (
@@ -52,7 +51,6 @@ def register_channel_callbacks(app: dash.Dash, app_state: AppState) -> None:
         if visible_data is None:
             return prev_selected or []
 
-        prev = set(prev_selected or [])
         visible_keys = {row["key"] for row in visible_data}
 
         # Keys currently checked in the visible table
@@ -77,7 +75,7 @@ def register_channel_callbacks(app: dash.Dash, app_state: AppState) -> None:
         Output("selected-channels-badges", "children"),
         [Input("selected-channels-store", "data")],
     )
-    def update_badges(selected_keys: list[str] | None) -> list:
+    def update_badges(selected_keys: list[str] | None) -> list[Any]:
         return build_selected_channels_badges(selected_keys or [], app_state)
 
     @app.callback(
@@ -104,7 +102,7 @@ def register_channel_callbacks(app: dash.Dash, app_state: AppState) -> None:
         direction: str | None,
         selected_keys: list[str] | None,
         all_rows: list[dict[str, Any]] | None,
-    ) -> tuple[list[dict[str, Any]], list[int], list[dict]]:
+    ) -> tuple[list[dict[str, Any]], list[int], list[dict[str, Any]]]:
         """Filter rows AND recompute selected_rows + color styling."""
         if not all_rows:
             return [], [], []
@@ -135,7 +133,7 @@ def register_channel_callbacks(app: dash.Dash, app_state: AppState) -> None:
                 selected_indices.append(idx)
 
         # Build style_data_conditional for coloring selected rows
-        style_cond: list[dict] = []
+        style_cond: list[dict[str, Any]] = []
         for idx, row in enumerate(filtered):
             if row["key"] in color_map:
                 ci = color_map[row["key"]]
@@ -168,7 +166,7 @@ def register_channel_callbacks(app: dash.Dash, app_state: AppState) -> None:
         Output("per-channel-overrides", "children"),
         [Input("selected-channels-store", "data")],
     )
-    def update_per_channel_overrides(selected_keys: list[str] | None) -> list:
+    def update_per_channel_overrides(selected_keys: list[str] | None) -> list[Any]:
         """Show per-channel CFC override controls for selected channels."""
         return build_per_channel_override_rows(
             selected_keys or [],

src/impakt/web/callbacks/corridor_callbacks.py

@@ -9,11 +9,9 @@ Handles:
 from __future__ import annotations
 
 import base64
-import io
 from typing import Any
 
 import dash
-import numpy as np
 from dash import Input, Output, State, html
 from dash.exceptions import PreventUpdate
 
@@ -42,7 +40,7 @@ def register_corridor_callbacks(app: dash.Dash, app_state: AppState) -> None:
         filename: str | None,
         corridor_name: str | None,
         current_corridors: list[dict[str, Any]] | None,
-    ) -> tuple[list[dict[str, Any]], Any, list]:
+    ) -> tuple[list[dict[str, Any]], Any, list[Any]]:
         if contents is None:
             raise PreventUpdate
 
@@ -123,7 +121,7 @@ def register_corridor_callbacks(app: dash.Dash, app_state: AppState) -> None:
             )
 
 
-def _build_corridor_list(corridors: list[dict[str, Any]]) -> list:
+def _build_corridor_list(corridors: list[dict[str, Any]]) -> list[Any]:
     """Build the active corridors display list."""
     if not corridors:
         return [html.Div("No corridors loaded", className="text-muted", style={"fontSize": "10px"})]

src/impakt/web/callbacks/criteria_callbacks.py

@@ -10,7 +10,6 @@ from typing import Any
 
 import dash
 from dash import Input, Output, State, html
-from dash.exceptions import PreventUpdate
 
 from impakt.web.components.criteria import build_criteria_results_display
 from impakt.web.state import AppState

src/impakt/web/callbacks/cursor_callbacks.py

@@ -10,8 +10,7 @@ from __future__ import annotations
 from typing import Any
 
 import dash
-from dash import Input, Output, State, html
-from dash.exceptions import PreventUpdate
+from dash import Input, Output, State
 
 from impakt.plot.engine import DEFAULT_COLORS
 from impakt.web.callbacks.plot_callbacks import _resolve_channels
@@ -61,7 +60,7 @@ def register_cursor_callbacks(app: dash.Dash, app_state: AppState) -> None:
         selected_keys: list[str] | None,
         cfc_value: str,
         y_align: bool,
-        channel_overrides: dict | None,
+        channel_overrides: dict[str, Any] | None,
         x_align_method: str,
         x_align_value: float | None,
         show_resultant: bool,
@@ -105,7 +104,7 @@ def register_cursor_callbacks(app: dash.Dash, app_state: AppState) -> None:
         x_align_method: str,
         x_align_value: float | None,
         show_resultant: bool,
-    ) -> list[dict]:
+    ) -> list[dict[str, Any]]:
         if not selected_keys:
             return []
 
@@ -119,7 +118,7 @@ def register_cursor_callbacks(app: dash.Dash, app_state: AppState) -> None:
             show_resultant,
         )
 
-        style_cond: list[dict] = []
+        style_cond: list[dict[str, Any]] = []
         for i in range(len(channels)):
             color = DEFAULT_COLORS[i % len(DEFAULT_COLORS)]
             style_cond.append(

src/impakt/web/callbacks/export_callbacks.py

@@ -8,14 +8,12 @@ Handles:
 
 from __future__ import annotations
 
-import io
-import tempfile
 from typing import Any
 
 import dash
 import numpy as np
 import pandas as pd
-from dash import Input, Output, State, dcc, html, no_update
+from dash import Input, Output, State, dcc, html
 from dash.exceptions import PreventUpdate
 
 from impakt.web.callbacks.plot_callbacks import _resolve_channels
@@ -116,7 +114,6 @@ def register_export_callbacks(app: dash.Dash, app_state: AppState) -> None:
 
         # Generate report
         try:
-            from impakt.report.engine import generate_protocol_report
             import tempfile
 
             with tempfile.NamedTemporaryFile(suffix=".html", delete=False, mode="w") as f:

src/impakt/web/callbacks/file_callbacks.py

@@ -13,7 +13,7 @@ from pathlib import Path
 from typing import Any
 
 import dash
-from dash import Input, Output, State, html, no_update
+from dash import Input, Output, State, no_update
 from dash.exceptions import PreventUpdate
 
 from impakt.web.state import AppState

src/impakt/web/callbacks/plot_callbacks.py

@@ -17,7 +17,6 @@ import dash
 import numpy as np
 import plotly.graph_objects as go
 from dash import Input, Output, State
-from dash.exceptions import PreventUpdate
 
 from impakt.channel.model import Channel
 from impakt.plot.engine import DEFAULT_COLORS, PlotEngine
@@ -148,7 +147,7 @@ def _build_plot_spec(
     channels: list[tuple[str, Channel]],
     cursor_x1: float | None,
     cursor_x2: float | None,
-    corridors: list[dict] | None = None,
+    corridors: list[dict[str, Any]] | None = None,
 ) -> PlotSpec:
     """Build a PlotSpec from resolved channels and UI state.
 
@@ -228,8 +227,8 @@ def register_plot_callbacks(app: dash.Dash, app_state: AppState) -> None:
         x_align_method: str,
         cursor_x1: float | None,
         cursor_x2: float | None,
-        channel_overrides: dict | None,
-        corridors_data: list[dict] | None,
+        channel_overrides: dict[str, Any] | None,
+        corridors_data: list[dict[str, Any]] | None,
         x_align_value: float | None,
     ) -> go.Figure:
         if not selected_keys:

src/impakt/web/components/channel_grid.py

@@ -14,13 +14,12 @@ without reordering.
 from __future__ import annotations
 
 import fnmatch
-import re
 from typing import Any
 
 import dash_bootstrap_components as dbc
 from dash import dash_table, dcc, html
 
-from impakt.channel.lookup import DIRECTIONS, MAIN_LOCATIONS, MEASUREMENTS
+from impakt.channel.lookup import MAIN_LOCATIONS, MEASUREMENTS
 from impakt.plot.engine import DEFAULT_COLORS
 from impakt.web.state import AppState
 
@@ -277,7 +276,7 @@ def build_channel_grid(app_state: AppState) -> dbc.Card:
 def build_selected_channels_badges(
     selected_keys: list[str],
     app_state: AppState,
-) -> list:
+) -> list[Any]:
     """Build badge pills showing currently selected channels."""
     if not selected_keys:
         return [

src/impakt/web/components/channel_values.py

@@ -14,9 +14,8 @@ from __future__ import annotations
 
 from typing import Any
 
-import numpy as np
-
 import dash_bootstrap_components as dbc
+import numpy as np
 from dash import dash_table, dcc, html
 
 from impakt.channel.model import Channel
@@ -152,7 +151,7 @@ def build_channel_values_panel() -> dbc.Card:
                             "whiteSpace": "nowrap",
                         },
                         style_cell_conditional=[
-                            # Fixed-width columns (percentages sum to ~74%, leaving ~26% for Description)
+                            # Fixed-width columns (~74%, leaving ~26% for Description)
                             {
                                 "if": {"column_id": "ch_num"},
                                 "width": "3%",

src/impakt/web/components/corridors.py

@@ -8,8 +8,6 @@ Provides:
 
 from __future__ import annotations
 
-from typing import Any
-
 import dash_bootstrap_components as dbc
 from dash import dcc, html
 

src/impakt/web/components/criteria.py

@@ -12,7 +12,7 @@ from typing import Any
 import dash_bootstrap_components as dbc
 from dash import html
 
-from impakt.channel.model import Channel, TestData
+from impakt.channel.model import TestData
 from impakt.criteria import (
     CriterionResult,
     chest_deflection,
@@ -21,7 +21,6 @@ from impakt.criteria import (
     hic15,
     nij,
     tibia_index,
-    viscous_criterion,
 )
 from impakt.protocol.base import Color, ProtocolResult, Rating
 
@@ -229,7 +228,7 @@ def build_criteria_results_display(
     protocol_result: ProtocolResult | None = None,
 ) -> html.Div:
     """Build the criteria results display."""
-    elements: list = []
+    elements: list[Any] = []
 
     # Protocol summary
     if protocol_result:

src/impakt/web/components/header.py

@@ -7,7 +7,7 @@ and template selector.
 from __future__ import annotations
 
 import dash_bootstrap_components as dbc
-from dash import dcc, html
+from dash import html
 
 from impakt.web.state import AppState
 

src/impakt/web/components/math_builder.py

@@ -8,10 +8,8 @@ Example: sqrt(ax**2 + az**2) with ax=Head Accel X, az=Head Accel Z
 
 from __future__ import annotations
 
-from typing import Any
-
 import dash_bootstrap_components as dbc
-from dash import dcc, html
+from dash import html
 
 from impakt.web.state import AppState
 

src/impakt/web/components/plot_grid.py

@@ -6,10 +6,11 @@ plot panes. Each pane has its own channel selection and axis labels.
 
 from __future__ import annotations
 
+from typing import Any
+
 import dash_bootstrap_components as dbc
 from dash import dcc, html
 
-
 # Layout presets: (rows, cols)
 LAYOUT_PRESETS: dict[str, tuple[int, int]] = {
     "1x1": (1, 1),
@@ -57,7 +58,7 @@ def build_plot_grid(layout: str = "1x1") -> html.Div:
     )
 
 
-def _build_panes(layout: str) -> list:
+def _build_panes(layout: str) -> list[Any]:
     """Build plot pane elements for a given layout."""
     rows, cols = LAYOUT_PRESETS.get(layout, (1, 1))
     pane_count = rows * cols
@@ -110,7 +111,7 @@ def _build_single_pane(pane_idx: int, total_panes: int) -> dbc.Card:
     )
 
 
-def build_empty_plot_figure() -> dict:
+def build_empty_plot_figure() -> dict[str, Any]:
     """Build an empty plot figure with instruction text."""
     return {
         "data": [],

src/impakt/web/components/templates.py

@@ -10,10 +10,8 @@ Provides:
 
 from __future__ import annotations
 
-from typing import Any
-
 import dash_bootstrap_components as dbc
-from dash import dcc, html
+from dash import html
 
 from impakt.web.state import AppState
 

src/impakt/web/components/transforms.py

@@ -5,6 +5,8 @@ Provides global transform defaults and per-channel override capability.
 
 from __future__ import annotations
 
+from typing import Any
+
 import dash_bootstrap_components as dbc
 from dash import dcc, html
 
@@ -92,7 +94,7 @@ def build_transform_panel() -> dbc.Card:
 def build_per_channel_override_rows(
     selected_keys: list[str],
     overrides: dict[str, dict[str, str]],
-) -> list:
+) -> list[Any]:
     """Build per-channel override controls for selected channels.
 
     Shows a compact row per selected channel with a CFC dropdown override.

src/impakt/web/layout.py

@@ -8,13 +8,12 @@ Two tabs:
 
 from __future__ import annotations
 
-from typing import Any
-
 import dash_bootstrap_components as dbc
 from dash import dcc, html
 
 from impakt.web.components.channel_grid import build_channel_grid
 from impakt.web.components.channel_values import build_channel_values_panel
+from impakt.web.components.corridors import build_corridor_panel
 from impakt.web.components.criteria import build_criteria_panel
 from impakt.web.components.header import (
     build_header,
@@ -22,9 +21,8 @@ from impakt.web.components.header import (
     build_overlay_modal,
     build_test_info_panel,
 )
-from impakt.web.components.plot_grid import build_plot_grid
-from impakt.web.components.corridors import build_corridor_panel
 from impakt.web.components.math_builder import build_math_panel
+from impakt.web.components.plot_grid import build_plot_grid
 from impakt.web.components.report import build_report_panel
 from impakt.web.components.templates import build_template_panel
 from impakt.web.components.transforms import build_transform_panel

src/impakt/web/state.py

@@ -15,12 +15,10 @@ import logging
 from pathlib import Path
 from typing import Any
 
-from impakt.channel.model import Channel, TestData
+from impakt.channel.model import Channel
 from impakt.script.api import Session
 from impakt.template.library import TemplateLibrary
 from impakt.template.model import PlotDefinition, TemplateSpec
-from impakt.transform.align import YAlign
-from impakt.transform.cfc import CFCFilter
 
 logger = logging.getLogger(__name__)